B BNG Cg43xasr9k

Cisco ASR 9000 Series Aggregation Services Router Broadband
Network Gateway Configuration Guide, Release 4.3.x

Americas Headquarters
Cisco Systems, Inc.
170 West Tasman Drive
San Jose, CA 95134-1706
USA
http://www.cisco.com
Tel: 408 526-4000
800 553-NETS (6387)
Fax: 408 527-0883
Text Part Number: OL-28375-03
Cisco Systems, Inc. All rights reserved.

C ONT E NT S
P r e f a c e Preface xi
Changes to This Document xi
Obtaining Documentation and Submitting a Service Request xi
C H A P T E R 1 New and Changed Feature Information in Cisco IOS XR Release 4.3.x 1
New and Changed Information 1
C H A P T E R 2 Broadband Network Gateway Overview 15
Understanding BNG 15
BNG Architecture 16
BNG Role in ISP Network Models 18
BNG Packaging 19
Installing and Activating the BNG Pie on Cisco ASR 9000 Series Router 19
BNG Configuration Process 21
Hardware Requirements for BNG 21
BNG Interoperability 22
C H A P T E R 3 Configuring Authentication, Authorization, and Accounting Functions 25
AAA Overview 25
Using RADIUS Server Group 27
Configuring RADIUS Server Group 27
Specifying Method List 29
Configuring Method Lists for AAA 30
Defining AAA Attributes 31
Creating Attributes of Specific Format 33
Configuring RADIUS Attribute List 37
Configuring RADIUS Attribute Format 39
Cisco ASR 9000 Series Aggregation Services Router Broadband Network Gateway Configuration Guide, Release
4.3.x
OL-28375-03 iii
Configuring RADIUS Attribute Nas-port-type 40
Configuring AAA Attribute Format Function 42
Making RADIUS Server Settings 43
Configuring RADIUS Server Settings 43
Configuring Automated Testing 47
Setting IP DSCP for RADIUS Server 49
Balancing Transaction Load on the RADIUS Server 50
Configuring Load Balancing for Global RADIUS Server Group 50
Configuring Load Balancing for a Named RADIUS Server Group 51
Throttling of RADIUS Records 53
Configuring RADIUS Throttling Globally 53
Configuring RADIUS Throttling on a Server Group 55
RADIUS Change of Authorization (CoA) Overview 56
Service Accounting 58
Configuring Service Accounting 60
Statistics Infrastructure 62
Configuring Statistics IDs (statsD) 62
Understanding Per-VRF AAA Function 63
RADIUS Double-Dip Feature 64
Additional References 64
C H A P T E R 4 Activating Control Policy 67
Control Policy Overview 67
Creating Class-Map 68
Configuring a Class-Map 69
Creating Policy-Map 70
Control Policy Events 71
Configuring a Policy-Map 72
Activating Policy-Map 74
Enabling a Service-Policy on a Subscriber Interface 74
Defining Dynamic Templates 75
C H A P T E R 5 Establishing Subscriber Sessions 79
Subscriber Session Overview 79
Cisco ASR 9000 Series Aggregation Services Router Broadband Network Gateway Configuration Guide,
Release 4.3.x
iv OL-28375-03
Contents
Establishing IPoE Session 81
Enabling IPv4 or IPv6 on an Access Interface 82
Creating Dynamic Template for IPv4 or IPv6 Subscriber Session 84
Creating a Policy-Map to Run During IPoE Session 87
Enabling IPoE Subscribers on an Access Interface 88
Establishing PPPoE Session 91
Provisioning PPP PTA Session 92
Creating PPPoE Profiles 93
Creating a PPP Dynamic-Template 94
Creating a Policy-Map to Run During PPPoE Session 95
Applying the PPPoE Configurations to an Access Interface 97
Provisioning PPP LAC Session 99
L2TP Reassembly on LAC 100
Enabling L2TP Reassembly on LAC 101
L2TP Access Concentrator Stateful Switchover 102
Enabling LAC SSO 102
Enabling RPFO on Process-failures 104
Configuring the VPDN Template 105
Configuring Maximum Simultaneous VPDN Sessions 107
Activating VPDN Logging 109
Configuring Options to Apply on Calling Station ID 110
Configuring L2TP Session-ID Commands 111
Configuring L2TP Class Options 113
Configuring Softshut for VPDN 115
PPPoE Smart Server Selection 116
Configuring PADO Delay 117
PPPoE Session Limit and Throttle 119
PPPoE Session Limit 119
Configuring PPPoE Session Limit 119
PPPoE Session Throttle 121
Configuring PPPoE Session Throttle 121
Making DHCP Settings 122
Enabling DHCP Proxy 123
Configuring DHCP IPv4 Profile Proxy Class 124
Configuring a Circuit-ID for an Interface 126
4.3.x
OL-28375-03 v
Contents
Configuring a Remote-ID 127
Configuring the Client Lease Time 128
Attaching a Proxy Profile to an Interface 130
Specifying DHCP Lease Limit 131
Specifying the Lease Limit for a Circuit-ID 132
Specifying the Lease Limit for a Remote-ID 133
Specifying the Lease Limit for an Interface 134
Understanding DHCP Option-82 135
Option 82 Relay Information Encapsulation 136
Configuring DHCPv4 Class of Service (CoS) 136
DHCPv6 Overview 136
DHCPv6 Server and DHCPv6 Proxy 137
Enabling DHCPv6 for Different Configuration Modes 138
Setting Up DHCPv6 Parameters 141
DHCPv6 Features 143
High Availability Support for DHCPv6 143
DHCPv6 Prefix Delegation 144
IPv6 IPoE Subscriber Support 144
Configuring IPv6 IPoE Subscriber Interface 144
IPv6 PPPoE Subscriber Support 152
Configuring IPv6 PPPoE Subscriber Interfaces 152
Ambiguous VLAN Support 159
Configuring Ambiguous VLANs 159
DHCPv6 Address or Prefix Pool 162
Configuring IPv6 Address or Prefix Pool Name 162
DHCPv6 Dual-Stack Lite Support 165
Configuring AFTR Fully Qualified Domain Name for DS-Lite 165
VRF Awareness in DHCPv6 167
Defining VRF in a Dynamic Template 167
Packet Handling on Subscriber Interfaces 168
IPv6 Neighbor Discovery 170
C H A P T E R 6 Deploying the Quality of Service (QoS) 173
Quality of Service Overview 173
Release 4.3.x
vi OL-28375-03
Contents
Configuring Service-policy and Applying Subscriber Settings Through RADIUS 174
Configuring Service-policy and Applying Subscriber Settings Through Dynamic Template 176
Parameterized QoS 178
Parameterized QoS Syntax 179
Configuring Parameterized QoS Policy Through RADIUS 184
Modifying Service Policy through CoA 187
QoS Accounting 189
Configuring QoS Accounting 190
Support for Shared Policy Instance 191
Configuring a Policy with SPI in the Input or Output Direction Using Dynamic Template 192
Configuring a Policy with SPI in the Input or Output Direction Using RADIUS 196
Merging QoS Policy-maps 198
Enabling Policy-maps Merge 199
QoS Features Supported on BNG 202
VLAN Policy on Access Interface 206
Configuring Policy on S-VLAN 207
Configuring VLAN Policy on an Access Interface 208
C H A P T E R 7 Configuring Subscriber Features 211
Excessive Punt Flow Trap 211
Enabling Excessive Punt Flow Trap Processing 213
Access Control List and Access Control List-based Forwarding 215
Configuring Access-Control Lists 215
Activating ACL 217
Support for Lawful Intercept 218
Per-session Lawful Intercept 219
Disabling SNMP-based Lawful Intercept 220
Configuring the Inband Management Plane Protection Feature 221
Enabling the Mediation Device to Intercept VoIP and Data Sessions 221
Radius-based Lawful Intercept 224
Enabling RADIUS-based Lawful Intercept 224
TCP MSS Adjustment 226
Configuring the TCP MSS Value of TCP Packets 227
Subscriber Session on Ambiguous VLANs 230
4.3.x
OL-28375-03 vii
Contents
Establishing Subscriber Session on Ambiguous VLANs 230
uRPF 232
Multicast Services 233
Multicast Coexistence 233
Enabling Address Family for the VRF 233
Multicast Replication 234
HQoS Correlation 234
Configuring Minimum Unicast Bandwidth 235
Configuring Multicast HQOS Correlation Mode or Passive Mode 236
IGMP to Unicast QoS Shaper Correlation 237
Configuring the IGMP to HQoS Correlation Feature in a VRF 238
Configuring route-policy for Unicast QoS Shaper 239
Configuring IGMP Parameters for Subscriber Interfaces 241
IGMP Accounting 243
Configuring IGMP Accounting 243
DAPS Support 245
Configuring IPv4 Distributed Address Pool Service 245
Creating a Configuration Pool Submode 246
Configuring the Subnet Number and Mask for an Address Pool 248
Specifying a Range of IPv6 Addresses 250
Specifying a Utilization Threshold 251
Specifying the Length of the Prefix 253
Specifying a Set of Addresses or Prefixes Inside a Subnet 254
HTTP Redirect Using PBR 256
Identifying HTTP Destinations for Redirection 257
Configuring Class Maps for HTTP Redirection 261
Configuring Policy Map for HTTP Redirect 263
Configuring Dynamic Template for Applying HTTPR Policy 266
Configuring Web Logon 267
A P P E ND I X A XML Support for BNG Features 273
AAA XML Support 273
DHCP XML Support 276
Control Policy XML Support 279
Release 4.3.x
viii OL-28375-03
Contents
DAPS XML Support 282
PPPoE XML Support 283
Subscriber Database XML Support 285
A P P E ND I X B RADIUS Attributes 291
RADIUS IETF Attributes 291
IETF Tagged Attributes on LAC 293
RADIUS Vendor-Specific Attributes 294
Vendor-Specific Attributes for Account Operations 298
RADIUS ADSL Attributes 298
RADIUS ASCEND Attributes 299
RADIUS Microsoft Attributes 299
RADIUS Disconnect-Cause Attributes 299
A P P E ND I X C Action Handlers 305
4.3.x
OL-28375-03 ix
Contents
Release 4.3.x
x OL-28375-03
Contents
Preface
This Preface contains these sections:
Changes to This Document, page xi
Obtaining Documentation and Submitting a Service Request, page xi
Changes to This Document
This table lists the technical changes made to this document since it was first released.
Table 1: Changes to This Document
Summary Date Revision
Republished with documentation updates for
Cisco IOS XR Release 4.3.2 features.
September 2013 OL-28375-03
Republished with documentation updates for
Cisco IOS XR Release 4.3.1 features.
May 2013 OL-28375-02
Initial release of this document. December 2012 OL-28375-01
Obtaining Documentation and Submitting a Service Request
For information on obtaining documentation, using the Cisco Bug Search Tool (BST), submitting a service
request, and gathering additional information, see What's New in Cisco Product Documentation, at: http://
www.cisco.com/en/US/docs/general/whatsnew/whatsnew.html.
Subscribe to What's New in Cisco Product Documentation, which lists all new and revised Cisco technical
documentation, as an RSS feed and deliver content directly to your desktop using a reader application. The
RSS feeds are a free service.
4.3.x
OL-28375-03 xi
Release 4.3.x
xii OL-28375-03
Preface
Obtaining Documentation and Submitting a Service Request
C HAP T E R 1
New and Changed Feature Information in Cisco
IOS XR Release 4.3.x
This table summarizes the new and changed feature information for the Cisco ASR 9000 Series Aggregation
Services Router Broadband Network Gateway Configuration Guide, and tells you where they are documented.
For a complete list of new and changed features in Cisco IOS XR Software, Release 4.3.x, see the New and
Changed Features in Cisco IOS XRSoftware, Release 4.3.x for Cisco ASR9000 Series Aggregation Services
Router document.
New and Changed Information, page 1
New and Changed Information
This table summarizes the new and changed feature information for the Cisco ASR 9000 Series Aggregation
Services Router Broadband Network Gateway Configuration Guide, and tells you where they are documented.
Table 2: New and Changed Features
Where Documented Introduced/Updated Release Description Feature
Establishing Subscriber
Sessions chapter:
Configuring DHCPv4
Class of Service (CoS),
on page 136
Refer BNG DHCP
Commands chapter in Cisco
ASR9000 Series Aggregation
Services Router Broadband
Network Gateway Command
Reference for information on
the commands used for
configuring DHCPv4 CoS.
Release 4.3.2 This feature was
introduced.
Configuring
DHCPv4 Class of
Service (CoS)
4.3.x
OL-28375-03 1
Configuring Subscriber
Features chapter:
Support for Lawful
Intercept, on page 218
Disabling SNMP-based
Lawful Intercept, on
page 220
Refer BNG AAA Commands
chapter in Cisco ASR 9000
Series Aggregation Services
Router Broadband Network
Gateway Command
lawful intercept.
Release 4.3.2 By default, Lawful
Intercept is no longer a part
of the Cisco IOS XR
software.
Lawful Intercept
(LI) as a separate
package
Sessions chapter:
PPPoE Smart Server
Selection, on page 116
Configuring PADO
Delay, on page 117
XML Support for BNG
Features chapter:
PPPoE XML Support,
on page 283
Refer PPPoE Commands
Gateway Command
configuring PADO delay for
PPPoE Smart Server
Selection feature.
introduced.
PPPoE Smart
Server Selection
Release 4.3.x
2 OL-28375-03
New and Changed Feature Information in Cisco IOS XR Release 4.3.x
Configuring Authentication,
Authorizing, and Accounting
Functions chapter:
NAS-Port-Type on
Interface or VLAN
Sub-interface, on page
34
Configuring RADIUS
Attribute
Nas-port-type, on page
40
Gateway Command
configuring NAS-Port-Type.
Release 4.3.1 NAS-Port-Type was made
configurable on interface or
VLAN sub-interface
NAS-Port-Type
on Interface or
VLAN
Sub-interface
Sessions chapter:
L2TP Reassembly on
LAC, on page 100
Enabling L2TP
Reassembly on LAC,
on page 101
Refer PPPoE LAC-Specific
configuring NAS-Port-Type.
introduced.
L2TPReassembly
on LAC
RADIUS Attributes Appendix
B chapter:
IETF Tagged Attributes
on LAC, on page 293
Release 4.3.1 The tag support was added
for IETF Attributes on
LAC.
IETF Tagged
Attributes on
LAC
4.3.x
OL-28375-03 3
Sessions chapter:
PPPoE Session Limit,
on page 119
PPPoE Session
Throttle, on page 121
Configuring PPPoE
Session Limit, on page
119
Configuring PPPoE
Session Throttle, on
page 121
Refer PPPoE Commands
Gateway Command
configuring PPPoE session
limit and throttle.
Release 4.3.1 The PPPoE session limit
and throttle support was
extended for more
parameters, with the
addition of a threshold for
the limit and request-count,
request-period and blocking
period for the throttle.
PPPoE Session
Limit and
Throttle
Sessions chapter:
Configuring
Ambiguous VLANs,
on page 159
Features chapter:
#unique_22
Release 4.3.1 Ambiguous VLAN
configuration was enhanced
to support two new
encapsulations.
Ambiguous
VLAN
Broadband Network Gateway
Overview chapter:
Hardware
Requirements for
BNG, on page 21
Release 4.3.1 BNG support is added on
Cisco ASR 9922 Series
Aggregation Services
Routers.
ASR 9922
Support
Release 4.3.x
4 OL-28375-03
Deploying the Quality of
Service (QoS) chapter:
VLAN Policy on
Access Interface, on
page 206
Configuring VLAN
Policy on an Access
Interface, on page 208
Refer QOS Commands
Gateway Command
configuring VLAN policy.
introduced.
VLAN Policy on
Access Interface
Sessions chapter:
Specifying DHCP
Lease Limit, on page
131
Release 4.3.1 Support for accepting
session limit per Circuit-ID
from AAA and overriding
the lease limit configured
from profile, was added.
Session Limit per
Circuit-ID
Sessions chapter:
L2TP Access
Concentrator Stateful
Switchover, on page
102
Enabling LAC SSO,
on page 102
Enabling RPFO on
Process-failures, on
page 104
Refer PPPoE LAC-Specific
configuring LAC SSO.
introduced.
LAC SSO
4.3.x
OL-28375-03 5
Configuring Authentication,
Authorizing, and Accounting
Functions chapter:
Service Accounting, on
page 58
Configuring Service
Accounting, on page
60
Statistics
Infrastructure, on page
62
Configuring Statistics
IDs (statsD), on page
62
Gateway Command
configuring Service
Accounting.
introduced.
Service
Accounting for
QoS
Deploying the Quality of
Service (QoS) chapter:
Merging QoS
Policy-maps, on page
198
Enabling Policy-maps
Merge, on page 199
Refer QOS Commands
Gateway Command
Merging QoS Policy-maps.
introduced.
Merging QoS
Policy-maps
Release 4.3.x
6 OL-28375-03
Sessions chapter:
Option 82 Relay
Information
Encapsulation, on page
136
Refer BNG DHCP
Option 82 Relay Information
Encapsulation.
introduced.
Option 82 Relay
Information
Encapsulation
4.3.x
OL-28375-03 7
Sessions chapter:
DHCPv6 Overview, on
page 136
DHCPv6 Server and
DHCPv6 Proxy, on
page 137
DHCPv6 Features, on
page 143
Enabling DHCPv6 for
Different Configuration
Modes, on page 138
Setting Up DHCPv6
Parameters, on page
141
Configuring IPv6 IPoE
Subscriber Interface,
on page 144
Configuring IPv6
PPPoE Subscriber
Interfaces, on page 152
Configuring IPv6
Address or Prefix Pool
Name, on page 162
Refer BNG DHCP
configuring DHCP IPv6.
Release 4.3.0 This feature was supported
on IPv6.
DHCPv6
Sessions chapter:
DHCPv6 Dual-Stack
Lite Support, on page
165
Configuring AFTR
Fully Qualified Domain
Name for DS-Lite, on
page 165
introduced.
DS-Lite
Release 4.3.x
8 OL-28375-03
Sessions chapter:
IPv6 Neighbor
Discovery, on page 170
Refer Neighbor Discovery
configuring IPv6 ND.
on IPv6.
IPv6 Neighbor
Discovery
Overview chapter:
BNG Packaging, on
page 19
Installing and
Activating the BNGPie
on Cisco ASR 9000
Series Router, on page
19
introduced.
BNG Packaging
Features chapter:
HTTP Redirect Using
PBR, on page 256
Identifying HTTP
Destinations for
Redirection, on page
257
Configuring Class
Maps for HTTP
Redirection, on page
261
Configuring Policy
Map for HTTP
Redirect, on page 263
Configuring Dynamic
Template for Applying
HTTPR Policy, on
page 266
on IPv6.
IPv6 Support on
HTTP-Redirect
4.3.x
OL-28375-03 9
Sessions chapter:
Establishing IPoE
Session, on page 81
Establishing PPPoE
Session, on page 91
uRPF, on page 232
Refer IPoE Commands and
PPPoE Commands chapters
in Cisco ASR 9000 Series
Aggregation Services Router
Command Reference for
information on the
commands used for
configuring IPv6 over IPoE
and PPPoE.
introduced.
IPv6 over PPPoE
and IPoE
Sessions and IPv6
uRPF for IPoE
and PTA
Overview chapter:
BNG Interoperability,
on page 22
introduced.
BNG Satellite
Enhancements
Release 4.3.x
10 OL-28375-03
Features chapter:
DAPS Support, on
page 245
Configuring the Subnet
Number and Mask for
an Address Pool, on
page 248
Specifying a Range of
IPv6 Addresses, on
page 250
Specifying a Utilization
Threshold, on page 251
Specifying the Length
of the Prefix, on page
253
Specifying a Set of
Addresses or Prefixes
Inside a Subnet, on
page 254
Refer Address Pool Service
configuring DAPS.
on IPv6.
IPv6 Support on
DAPS
Sessions chapter:
Packet Handling on
Subscriber Interfaces,
on page 168
with enhancements.
Packet Handling
on Subscriber
Interfaces
Enhancements
4.3.x
OL-28375-03 11
Deploying QoS chapter:
Support for Shared
Policy Instance, on
page 191
Configuring
Service-policy and
Applying Subscriber
Settings Through
Dynamic Template, on
page 176
Configuring a Policy
with SPI in the Input or
Output Direction Using
RADIUS, on page 196
Refer QoS Commands
Gateway Command
configuring QoS.
introduced.
QoS - Support for
Shared Policy
Instance
Features chapter:
Access Control List and
Access Control
List-based Forwarding,
on page 215
Configuring
Access-Control Lists,
on page 215
Activating ACL, on
page 217
Refer ACL and ABF
configuring QoS.
on IPv6.
IPv6 Support on
ACL
Release 4.3.x
12 OL-28375-03
Features chapter:
Excessive Punt Flow
Trap, on page 211
Enabling Excessive
Punt Flow Trap
Processing, on page 213
Refer Excessive Punt Flow
Trap Commands chapter in
Cisco ASR 9000 Series
Aggregation Services Router
Command Reference for
information on the
commands used for
configuring Excessive Punt
Flow Trap.
introduced.
Excessive Punt
Flow Trap
Features chapter:
TCP MSS Adjustment,
on page 226
Configuring the TCP
MSS Value of TCP
Packets, on page 227
introduced.
TCP MSS
Handling
Features chapter:
Radius-based Lawful
Enabling
RADIUS-based Lawful
introduced.
Radius-based
Lawful Intercept
4.3.x
OL-28375-03 13
Release 4.3.x
14 OL-28375-03
C HAP T E R 2
Broadband Network Gateway Overview
This chapter provides an overview of the Broadband Network Gateway (BNG) functionality implemented
on the Cisco ASR 9000 Series Router.
Understanding BNG, page 15
BNG Architecture, page 16
BNG Role in ISP Network Models, page 18
BNG Packaging, page 19
BNG Configuration Process, page 21
Hardware Requirements for BNG, page 21
BNG Interoperability, page 22
Understanding BNG
Broadband Network Gateway (BNG) is the access point for subscribers, through which they connect to the
broadband network. When a connection is established between BNGand Customer Premise Equipment (CPE),
the subscriber can access the broadband services provided by the Network Service Provide (NSP) or Internet
Service Provider (ISP).
BNG establishes and manages subscriber sessions. When a session is active, BNG aggregates traffic from
various subscriber sessions from an access network, and routes it to the network of the service provider.
BNG is deployed by the service provider and is present at the first aggregation point in the network, such as
the edge router. An edge router, like the Cisco ASR 9000 Series Router, needs to be configured to act as the
BNG. Because the subscriber directly connects to the edge router, BNGeffectively manages subscriber access,
and subscriber management functions such as:
Authentication, authorization and accounting of subscriber sessions
Address assignment
Security
Policy management
Quality of Service (QoS)
4.3.x
OL-28375-03 15
Some benefits of using BNG are:
The BNG router not only performs the routing function but also communicates with authentication,
authorization, and accounting (AAA) server to perform session management and billing functions. This
makes the BNG solution more comprehensive.
Different subscribers can be provided different network services. This enables the service provider to
customize the broadband package for each customer based on their needs.
BNG Architecture
The goal of the BNG architecture is to enable the BNG router to interact with peripheral devices (like CPE)
and servers (like AAA and DHCP), in order to provide broadband connectivity to subscribers and manage
subscriber sessions. The basic BNG architecture is shown in the following figure.
Figure 1: BNG Architecture
The BNG architecture is designed to perform these tasks:
Connecting with the Customer Premise Equipment (CPE) that needs to be served broadband services.
Establishing subscriber sessions using IPoE or PPPoE protocols.
Interacting with the AAA server that authenticates subscribers, and keeps an account of subscriber
sessions.
Interacting with the DHCP server to provide IP address to clients.
The four BNG tasks are briefly explained in the following sections.
Connecting with the CPE
BNG connects to the CPE through a multiplexor and Home Gateway (HG). The CPE represents the triple
play service in telecommunications, namely, voice (phone), video (set top box), and data (PC). The individual
Release 4.3.x
16 OL-28375-03
BNG Architecture
subscriber devices connect to the HG. In this example, the subscriber connects to the network over a Digital
Subscriber Line (DSL) connection. Therefore, the HG connects into a DSL Access Multiplexor (DSLAM).
Multiple HGs can connect to a single DSLAM that sends the aggregated traffic to the BNG router. The BNG
router routes traffic between the broadband remote access devices (like DSLAM or Ethernet Aggregation
Switch) and the service provider network.
Establishing Subscriber Sessions
Each subscriber (or more specifically, an application running on the CPE) connects to the network by a logical
session. Based on the protocol used, subscriber sessions are classified into two types:
PPPoE subscriber sessionThe PPP over Ethernet (PPPoE) subscriber session is established using the
point-to-point (PPP) protocol that runs between the CPE and BNG.
IPoE subscriber sessionThe IP over Ethernet (IPoE) subscriber session is established using IP protocol
that runs between the CPE and BNG; IP addressing is done using the DHCP protocol.
Interacting with the RADIUS Server
BNGrelies on an external Remote Authentication Dial-In User Service (RADIUS) server to provide subscriber
Authentication, Authorization, and Accounting (AAA) functions. During the AAAprocess, BNGuses RADIUS
to:
authenticate a subscriber before establishing a subscriber session
authorize the subscriber to access specific network services or resources
track usage of broadband services for accounting or billing
The RADIUS server contains a complete database of all subscribers of a service provider, and provides
subscriber data updates to the BNG in the form of attributes within RADIUS messages. BNG, on the other
hand, provides session usage (accounting) information to the RADIUS server. For more information about
RADIUS attributes, see RADIUS Attributes, on page 291.
BNG supports connections with more than one RADIUS server to have fail over redundancy in the AAA
process. For example, if RADIUS server A is active, then BNG directs all messages to the RADIUS server
A. If the communication with RADIUS server A is lost, BNG redirects all messages to RADIUS server B.
During interactions between the BNG and RADIUS servers, BNG performs load balancing in a round-robin
manner. During the load balancing process, BNG sends AAA processing requests to RADIUS server A only
if it has the bandwidth to do the processing. Else, the request is send to RADIUS server B.
Interacting with the DHCP Server
BNG relies on an external Dynamic Host Configuration Protocol (DHCP) server for address allocation and
client configuration functions. BNG can connect to more than one DHCP server to have fail over redundancy
in the addressing process. The DHCP server contains an IP address pool, from which it allocates addresses
to the CPE.
During the interaction between BNG and the DHCP server, BNG acts as a DHCP relay or DHCP proxy.
As the DHCP relay, BNG receives DHCP broadcasts from the client CPE, and forwards the request to the
DHCP server.
As the DHCP proxy, BNG itself maintains the address pool by acquiring it from DHCP server, and also
manages the IP address lease. BNG communicates on Layer 2 with the client Home Gateway, and on Layer
3 with the DHCP server.
4.3.x
OL-28375-03 17
BNG Architecture
The DSLAM modifies the DHCP packets by inserting subscriber identification information. BNG uses the
identification information inserted by the DSLAM, as well as the address assigned by the DHCP server, to
identify the subscriber on the network, and monitor the IP address lease.
BNG Role in ISP Network Models
The role of BNG is to pass traffic from the subscriber to the ISP. The manner in which BNG connects to the
ISP depends on the model of the network in which it is present. There are two types of network models:
Network Service Provider, on page 18
Access Network Provider, on page 19
Network Service Provider
The following figure shows the topology of a Network Service Provider model.
Figure 2: Network Service Provider Model
In the Network Service Provider model, the ISP (also called the retailer) directly provides the broadband
connection to the subscriber. As shown in the above figure, BNGis at the edge router, and its role is to connect
to the core network through uplinks.
Release 4.3.x
18 OL-28375-03
BNG Role in ISP Network Models
Access Network Provider
The following figure shows the topology of a Access Network Provider model.
Figure 3: Access Network Provider Model
In the Access Network Provider model, a network carrier (also called the wholesaler) owns the edge network
infrastructure, and provides the broadband connection to the subscriber. However, the network carrier does
not own the broadband network. Instead, the network carrier connects to one of the ISPs that manage the
broadband network.
BNG is implemented by the network carrier and its role is to hand the subscriber traffic off to one of several
ISPs. The hand-off task, from the carrier to the ISP, is implemented by Layer 2 Tunneling Protocol (L2TP)
or Layer 3 Virtual Private Networking (VPN). L2TP requires two distinct network components:
L2TP Access Concentrator (LAC)The LAC is provided by the BNG.
L2TP Network Server (LNS)The LNS is provided by the ISP.
BNG Packaging
The BNGpie, asr9k-bng-px.pie can be installed and activated on the Cisco ASR9000 Series Router to access
the BNGfeatures. The install, uninstall, activate and deactivate operations can be performed without rebooting
the router.
It is recommended that the relevant BNG configurations be removed from the running configuration of the
router, before uninstalling or deactivating the BNG pie.
Installing and Activating the BNG Pie on Cisco ASR 9000 Series Router
Perform this task to install and activate the BNG pie on the Cisco ASR 9000 Series Router:
4.3.x
OL-28375-03 19
BNG Packaging
SUMMARY STEPS
1. admin
2. install add {pie_location | source | tar}
3. install activate {pie_name | id}
DETAILED STEPS
Purpose Command or Action
Enters the administration mode. admin
Example:
RP/0/RSP0/CPU0:router# admin
Step 1
Installs the pie fromthe tftp location, on to the Cisco
ASR 9000 Series Router.
install add {pie_location | source | tar}
Example:
RP/0/RSP0/CPU0:router(admin)# install add
tftp://223.255.254.254/softdir/asr9k-bng-px.pie
Step 2
Activates the installed pie on the Cisco ASR 9000
Series Router.
install activate {pie_name | id}
Example:
RP/0/RSP0/CPU0:router(admin)# install activate
asr9k-bng-px.pie
Step 3
What to Do Next
During upgrade from Release 4.2.1 to Release 4.3.0, it is recommended that the Cisco ASR 9000 base
image pie (asr9k-mini-px.pie) is installed prior to installing the BNG pie (asr9k-bng-px.pie).
Note
After BNG pie is installed, you must copy BNG related configurations from the flash or tftp location to the
router. If BNGpie is deactivated and activated again, then load the removed BNGconfigurations by executing
the load configuration removed command from the configuration terminal.
Most of the BNGfeature configurations are moved to a newnamespace partition, and hence BNGfeatures
are not available by default now. To avoid inconsistent BNG configurations before, or after installing the
BNG pie, run the clear configuration inconsistency command, in EXEC mode.
Note
Release 4.3.x
20 OL-28375-03
Installing and Activating the BNG Pie on Cisco ASR 9000 Series Router
BNG Configuration Process
Configuring BNG on the Cisco ASR 9000 Series Router involves these stages:
Configuring RADIUS ServerBNGis configured to interact with the RADIUS server for authentication,
authorization, and accounting functions. For details, see Configuring Authentication, Authorization, and
Accounting Functions, on page 25.
Activating Control PolicyControl policies are activated to determine the action that BNG takes when
specific events occur. The instructions for the action are provided in a policy map. For details, see
Activating Control Policy, on page 67.
Establishing Subscriber SessionsConfigurations are done to set up one or more logical sessions, from
the subscriber to the network, for accessing broadband services. Each session is uniquely tracked and
managed. For details, see Establishing Subscriber Sessions, on page 79.
Deploying QoSQuality of Service (QoS) is deployed to provide control over a variety of network
applications and traffic types. For example, the service provider can have control over resources (example
bandwidth) allocated to each subscriber, provide customized services, and give priority to traffic belonging
to mission-critical applications. For details, see Deploying the Quality of Service (QoS), on page 173.
Configuring Subscriber FeaturesConfigurations are done to activate certain subscriber features that
provide additional capabilities like policy based routing, access control using access list and access
groups, and multicast services. For details, see Configuring Subscriber Features, on page 211.
Verifying Session EstablishmentEstablished sessions are verified and monitored to ensure that
connections are always available for use. The verification is primarily done using "show" commands.
Refer to the Cisco ASR9000 Series Aggregation Services Router Broadband Network Gateway Command
Reference guide for the list of various "show" commands.
To use a BNG command, you must be in a user group associated with a task group that includes the proper
task IDs. The Cisco ASR 9000 Series Aggregation Services Router Broadband Network Gateway Command
Reference guide includes the task IDs required for each command. If you suspect that the user group assignment
is preventing you from using a command, contact your AAA administrator for assistance.
Restriction
The Select VRF Download (SVD) must be disabled, when BNG is configured. For more information about
SVD, see the Cisco IOS XR Routing Configuration Guide for the Cisco XR 12000 Series Router.
Hardware Requirements for BNG
These hardwares support BNG:
BNG is supported on Satellite Network Virtualization (nV) system.
BNG is supported on Cisco ASR 9922 Series Aggregation Services Routers.
BNG is supported on Cisco ASR 9000 Series Aggregation Services Routers only with RSP-440 route
switch processors. The RSP 2 route switch processor does not support BNG.
4.3.x
OL-28375-03 21
BNG Configuration Process
Table 3: Line Cards and Modular Port Adapters Supported on BNG
Modular Port Adapters Line Cards
A9K-24X10GE-SE 24-Port 10-Gigabit Ethernet Line Card, Service
Edge Optimized
A9K-36X10GE-SE 36-Port 10-Gigabit Ethernet Line Card, Service
Edge Optimized
A9K-MOD80-SE 80 Gigabyte Modular Line Card, Service Edge
Optimized
A9K-MOD160-SE 160 Gigabyte Modular Line Card, Service Edge
Optimized
A9K-MPA-20GE 20-Port Gigabit Ethernet Modular Port Adapter
(MPA)
A9K-MPA-2X10GE 2-port 10-Gigabit Ethernet Modular Port Adapter
(MPA)
A9K-MPA-4X10GE 4-Port 10-Gigabit Ethernet Modular Port Adapter
(MPA)
A9K-MPA-2X40GE 2-port 40-Gigabit Ethernet Modular Port Adapter
(MPA)
A9K-MPA-1X40GE 1-Port 40-Gigabit Ethernet Modular Port Adapter
(MPA)
BNG Interoperability
The BNG interoperability allows BNG to exchange and use information with other larger heterogeneous
networks. These are the key features:
BNG Coexists with ASR9001
ASR9001 is a standalone high processing capability router that comprises of a route switch processor
(RSP), linecards (LC), and ethernet plugs (EPs). All BNG features are fully supported on the ASR9001
chassis.
BNG Supports nV Satellite
Two different topologies that nV Satellite supports are:
Bundled Ethernet ports on the CPE side of the Satellite node connected to the ASR9K through
non-bundle configuration (static-pinning).
Release 4.3.x
22 OL-28375-03
Non-bundle ports fromthe satellite out to the access network, and bundle ports between the Satellite
node and the ASR9K.
BNG interoperates with Carrier Grade NAT (CGN)
To address the impending threat fromIPv4 address space depletion, it is recommended that the remaining
or available IPv4 addresses be shared among larger numbers of customers. This is done by using CGN,
which primarily pulls the address allocation to a more centralized NAT in the service provider network.
NAT44 is a technology that uses CGN and helps manage depletion issues of the IPv4 address space.
BNG supports the ability to perform NAT44 translation on IPoE and PPPoE-based BNG subscriber
sessions.
Restrictions
This topology is supported on nV Satellite, but not supported on BNG:
Single Ethernet ports (non-bundle) on the CPE side of the Satellite node, connected to the ASR9K
through non-bundle configuration (static-pinning).
This topology is not supported on nV Satellite:
Bundled Ethernet ports on the CPE side of the Satellite node, connected to the ASR9K through
bundle Ethernet connections.
4.3.x
OL-28375-03 23
Release 4.3.x
24 OL-28375-03
C HAP T E R 3
Configuring Authentication, Authorization, and
Accounting Functions
This chapter provides information about configuring authentication, authorization, and accounting (AAA)
functions on the BNG router. BNG interacts with the RADIUS server to perform AAA functions. A group
of RADIUS servers form a server group that is assigned specific AAA tasks. A method list defined on a
server or server group lists methods by which authorization is performed. Some of the RADIUS features
include creating specific AAA attribute formats, load balancing of RADIUS servers, throttling of RADIUS
records, Change of Authorization (CoA), and Service Accounting for QoS. This chapter covers these topics:
AAA Overview, page 25
Using RADIUS Server Group, page 27
Specifying Method List, page 29
Defining AAA Attributes, page 31
Making RADIUS Server Settings, page 43
Balancing Transaction Load on the RADIUS Server, page 50
Throttling of RADIUS Records, page 53
RADIUS Change of Authorization (CoA) Overview, page 56
Service Accounting, page 58
Understanding Per-VRF AAA Function, page 63
Additional References, page 64
AAA Overview
AAA acts as a framework for effective network management and security. It helps in managing network
resources, enforcing policies, auditing network usage, and providing bill-related information. BNG connects
to an external RADIUS server that provides the AAA functions.
The RADIUS server performs the three independent security functions (authentication, authorization, and
accounting) to secure networks against unauthorized access. The RADIUS server runs the Remote
Authentication Dial-In User Service (RADIUS) protocol. (For details about RADIUS protocol, refer to RFC
4.3.x
OL-28375-03 25
2865). The RADIUS server manages the AAAprocess by interacting with BNG, and databases and directories
containing user information.
The RADIUS protocol runs on a distributed client-server system. The RADIUS client runs on BNG (Cisco
ASR 9000 Series Router) that sends authentication requests to a central RADIUS server. The RADIUS server
contains all user authentication and network service access information.
The AAA processes, the role of RADIUS server during these processes, and some BNG restrictions, are
explained in these sections:
Authentication
The authentication process identifies a subscriber on the network, before granting access to the network and
network services. The process of authentication works on a unique set of criteria that each subscriber has for
gaining access to the network. Typically, the RADIUS server performs authentication by matching the
credentials (user name and password) the subscriber enters with those present in the database for that subscriber.
If the credentials match, the subscriber is granted access to the network. Otherwise, the authentication process
fails, and network access is denied.
Authorization
After the authentication process, the subscriber is authorized for performing certain activity. Authorization is
the process that determines what type of activities, resources, or services a subscriber is permitted to use. For
example, after logging into the network, the subscriber may try to access a database, or a restricted website.
The authorization process determines whether the subscriber has the authority to access these network resources.
AAA authorization works by assembling a set of attributes based on the authentication credentials provided
by the subscriber. The RADIUS server compares these attributes, for a given username, with information
contained in a database. The result is returned to BNG to determine the actual capabilities and restrictions
that are to be applied for that subscriber.
Accounting
The accounting keeps track of resources used by the subscriber during network access. Accounting is used
for billing, trend analysis, tracking resource utilization, and capacity planning activities. During the accounting
process, a log is maintained for network usage statistics. The information monitored include, but are not
limited to - subscriber identities, applied configurations on the subscriber, the start and stop times of network
connections, and the number of packets and bytes transferred to, and from, the network.
BNG reports subscriber activity to the RADIUS server in the form of accounting records. Each accounting
record comprises of an accounting attribute value. This value is analyzed and used by the RADIUS server for
network management, client billing, auditing, etc.
The accounting records of the subscriber sessions may timeout if the BNGdoes not receive acknowledgments
from the RADIUS server. This timeout can be due to RADIUS server being unreachable or due to network
connectivity issues leading to slow performance of the RADIUS server. If the sessions on the BNG are not
acknowledged for their Account-Start request, loss of sessions on route processor fail over (RPFO) and other
critical failures are reported. It is therefore recommended that a RADIUS server deadtime be configured on
the BNG, to avoid loss of sessions. Once this value is configured, and if a particular session is not receiving
an accounting response even after retries, then that particular RADIUS server is considered to be non-working
and further requests are not sent to that server.
The radius-server deadtime limit command can be used to configure the deadtime for RADIUS server. For
details, see Configuring RADIUS Server Settings, on page 43.
Release 4.3.x
26 OL-28375-03
Configuring Authentication, Authorization, and Accounting Functions
AAA Overview
Restrictions
On BNG, local authentication and local authorization are not supported. It must be done by the RADIUS
server.
On session disconnect, transmission of the Accounting-Stop request to RADIUS may be delayed for a
few seconds while the system waits for the "final" session statistics to be collected from the hardware.
The Event-Timestamp attribute in that Accounting-Stop request should, however, reflect the time the
client disconnects, and not the transmission time.
Using RADIUS Server Group
A RADIUS server group is a named group of one or more RADIUS servers. Each server group is used for a
particular service. For example, in an AAA network configuration having two RADIUS server groups, the
first server group can be assigned the authentication and authorization task, while the second group can be
assigned the accounting task.
Server groups can include multiple host entries for the same server. Each entry, however, must have a unique
identifier. This unique identifier is created by combining an IP address and a UDP port number. Different
ports of the server, therefore, can be separately defined as individual RADIUS hosts providing a specific AAA
service. In other words, this unique identifier enables RADIUS requests to be sent to different UDP ports on
the same server. Further, if two different host entries on the same RADIUS server are configured for the same
service (like the authentication process), then the second host entry acts as a fail-over backup for the first one.
That is, if the first host entry fails to provide authentication services, BNG tries with the second host entry.
(The RADIUS host entries are tried in the order in which they are created.)
For assigning specific actions to the server group, see Configuring RADIUS Server Group, on page 27.
Configuring RADIUS Server Group
Perform this task to define a named server group as the server host.
SUMMARY STEPS
1. configure
2. aaa group server radius name
3. accounting accept radius_attribute_list_name
4. authorization reply accept radius_attribute_list_name
5. deadtime limit
6. load-balance method least-outstanding batch-size size ignore-preferred-server
7. server host_name acct-port accounting_port_number auth-port authentication_port_number
8. source-interface name value
9. vrf name
10. Use the commit or end command.
4.3.x
OL-28375-03 27
Using RADIUS Server Group
DETAILED STEPS
Enters global configuration mode. configure
Example:
RP/0/RSP0/CPU0:router# configure
Step 1
Configures the RADIUS server group named r1. aaa group server radius name
Example:
RP/0/RSP0/CPU0:router(config)# aaa group server radius r1
Step 2
Configures the radius attribute filter for the
accounting process to accept only the attributes
specified in the list.
accounting accept radius_attribute_list_name
Example:
RP/0/RSP0/CPU0:router(config-sg-radius)# accounting accept
att_list
Step 3
Configures the radius attribute filter for the
authorization process to accept only the attributes
specified in the list.
authorization reply accept radius_attribute_list_name
Example:
RP/0/RSP0/CPU0:router(config-sg-radius)# authorization
reply accept att_list1
Step 4
Configures the RADIUS server-group deadtime.
The deadtime limit is configured in minutes. The
range is from 1 to 1440, and the default is 0.
deadtime limit
Example:
RP/0/RSP0/CPU0:router(config-sg-radius)# deadtime 40
Step 5
Configures load balancing batch size after which
the next host is picked.
load-balance method least-outstanding batch-size size
ignore-preferred-server
Example:
RP/0/RSP0/CPU0:router(config-sg-radius)# load-balance
Step 6
method least-outstanding batch-size 50
ignore-preferred-server
Specifies the radius server and its host name.
Configures the UDP port for RADIUS accounting
server host_name acct-port accounting_port_number auth-port
authentication_port_number
Step 7
and authentication requests. The accounting and
Example:
RP/0/RSP0/CPU0:router(config-sg-radius)# server 1.2.3.4
acct-port 455 auth-port 567
authentication port number ranges from0 to 65535.
If no value is specified, then the default is 1645 for
auth-port, and 1646 for acct-port.
Release 4.3.x
28 OL-28375-03
Configuring RADIUS Server Group
Configures the RADIUS server-group
source-interface name and value for Bundle-Ether.
source-interface name value
Example:
RP/0/RSP0/CPU0:router(config-sg-radius)# source-interface
Bundle-Ether 455
Step 8
Configures the vrf to which the server radius group
belongs.
vrf name
Example:
RP/0/RSP0/CPU0:router(config-sg-radius)# vrf vrf_1
Step 9
commitSaves the configuration changes, and
remains within the configuration session.
Use the commit or end command. Step 10
endPrompts user to take one of these actions:
Yes Saves configuration changes and exits
the configuration session.
NoExits the configuration session without
committing the configuration changes.
CancelRemains in the configuration mode,
without committing the configuration changes.
Configuring Radius Server-Group: An example
configure
aaa group server radius r1
accounting accept r1 r2
authorization reply accept a1 a2
deadtime 8
load-balance method least-outstanding batch-size 45 ignore-preferred-server
server host_name acct-port 355 auth-port 544
source-interface Bundle-Ether100.10
vrf vrf_1
!
end
Specifying Method List
Method lists for AAA define the methods using which authorization is performed, and the sequence in which
these methods are executed. Before any defined authentication method is performed, the method list must be
applied to the configuration mechanismresponsible for validating user-access credentials. The only exception
to this requirement is the default method list (named "default"). The default method list is automatically
applied if no other method list is defined. A defined method list overrides the default method list.
On BNG, you have to specify the method list and the server group that will be used for AAA services. For
specifying method lists, see Configuring Method Lists for AAA, on page 30.
4.3.x
OL-28375-03 29
Specifying Method List
Configuring Method Lists for AAA
Perform this task to assign the method list to be used by the server group for subscriber authentication,
authorization, and accounting.
SUMMARY STEPS
1. configure
2. aaa authentication subscriber default method-list-name group server-group-name
3. aaa authorization subscriber default method-list-name group server-group-name |radius
4. aaa accounting subscriber default method-list-name group server-group-name
DETAILED STEPS
Example:
Step 1
Configures the method-list which will be applied by default for
subscriber authentication. You can either enter 'default' or a
aaa authentication subscriber default
method-list-name group server-group-name
Step 2
user-defined name for the AAA method-list. Also, enter the name
of the server group, on which the method list is applied.
Example:
RP/0/RSP0/CPU0:router(config)# aaa
authentication subscriber default method1
group group1 radius group group2 group
group3 ...
subscriber authorization. You can either enter 'default' or a
aaa authorization subscriber default
method-list-name group server-group-name |radius
Step 3
user-defined name for the AAA method-list. Also, enter the name
of the server group, on which the method list is applied.
Example:
authorization subscriber default method1
group group1 radius group group2 group
group3 ...
subscriber accounting. You can either enter 'default' or a user-defined
aaa accounting subscriber default
method-list-name group server-group-name
Step 4
name for the AAA method-list. Also, enter the name of the server
group, on which the method list is applied.
Example:
accounting subscriber default method1 group
group1 radius group group2 group group3 ...
commitSaves the configuration changes, and remains within the
configuration session.
Release 4.3.x
30 OL-28375-03
Configuring Method Lists for AAA
YesSaves configuration changes and exits the configuration
session.
NoExits the configuration session without committing the
configuration changes.
CancelRemains in the configuration mode, without
Configuring Method-list for AAA: An example
configure
aaa authentication subscriber default group radius group rad2 group rad3..
aaa authorization subscriber default group radius group rad1 group rad2 group rad3..
aaa accounting subscriber default group radius group rad1 group rad2 group rad3..
!
!
end
Defining AAA Attributes
The AAA attribute is an element of RAIDUS packet. A RAIDUS packet transfers data between a RADIUS
server and a RADIUS client. The AAA attribute parameter, and its value - form a Attribute Value Pair (AVP).
The AVP carries data for both requests and responses for the AAA transaction.
The AAAattributes either can be predefined as in Internet Engineering Task Force (IETF) attributes or vendor
defined as in vendor-specific attributes (VSAs). For more information about the list of BNG supported
attributes, see RADIUS Attributes, on page 291.
The RADIUS server provides configuration updates to BNG in the form of attributes in RADIUS messages.
The configuration updates can be applied on a subscriber during session setup through two typical methods
per-user attributes, which applies configuration on a subscriber as part of the subscriber's authentication Access
Accept, or through explicit domain, port, or service authorization Access Accepts. This is all controlled by
the Policy Rule Engine's configuration on the subscriber.
When BNG sends an authentication or an authorization request to an external RADIUS server as an Access
Request, the server sends back configuration updates to BNG as part of the Access Accept. In addition to
RADIUS configuring a subscriber during setup, the server can send a change of authorization (CoA) message
autonomously to the BNG during the subscriber's active session lifecycle, even when the BNG did not send
a request. These RADIUS CoA updates act as dynamic updates, referencing configured elements in the BNG
and instructing the BNG to update a particular control policy or service policy.
BNG supports the concept of a "service", which is a group of configured features acting together to represent
that service. Services can be represented as either features configured on dynamic-templates through CLI, or
as features configured as RADIUS attributes inside Radius Servers. Services are activated either directly from
CLI or RADIUS through configured "activate" actions on the Policy Rule Engine, or through CoA
"activate-service" requests. Services can also be deactivated directly (removing all the involved features within
4.3.x
OL-28375-03 31
the named service) through configured "deactivate" action on the Policy Rule Engine or through CoA
"deactivate-service" requests.
The attribute values received from RADIUS interact with the subscriber session in this way:
BNG merges the values received in the RADIUS update with the existing values that were provisioned
statically by means of CLI commands, or from prior RADIUS updates.
In all cases, values received in a RADIUS update take precedence over any corresponding CLI provisioned
values or prior RADIUS updates. Even if you reconfigured the CLI provisioned values, the system does
not override session attributes or features that were received in a RADIUS update.
Changes made to CLI provision values on the dynamic template take effect immediately on all sessions
using that template, assuming the template features have not already been overridden by RADIUS. Same
applies to service updates made through CoA "service-update" requests.
AAA Attribute List
An attribute list is named list that contains a set of attributes. You can configure the RADIUS server to use a
particular attribute list to perform the AAA function.
To create an attribute list, see Configuring RADIUS Attribute List, on page 37.
AAA Attribute Format
It is possible to define a customized format for some attributes. The configuration syntax for creating a new
format is:
aaa attribute format <format-name> format-string [length] <string> *[<Identity-Attribute>]
where:
format-name Specifies the name given to the attribute format. This name is referred when the format
is applied on an attribute.
length (Optional) Specifies the maximum length of the formatted attribute string. If the final length
of the attribute string is greater than the value specified in LENGTH, it is truncated to LENGTH bytes.
The maximum value allowed for LENGTH is 255. If the argument is not configured, the default is also
255.
string Contains regular ASCII characters that includes conversion specifiers. Only the % symbol is
allowed as a conversion specifier in the STRING. The STRING value is enclosed in double quotes.
Identity-Attribute Identifies a session, and includes user-name, ip-address, and mac-address. A list
of currently-defined identity attributes is displayed on the CLI.
Once the format is defined, the FORMAT-NAME can be applied to various AAAattributes such as username,
nas-port-ID, calling-station-ID, and called-station-ID. The configurable AAA attributes that use the format
capability are explained in the section Creating Attributes of Specific Format, on page 33.
To create a customized nas-port attribute and apply a predefined format to nas-port-IDattribute , see Configuring
RADIUS Attribute Format, on page 39.
Specific functions can be defined for an attribute format for specific purposes. For example, if the input
username is "text@abc.com", and only the portion after "@" is required as the username, a function can be
defined to retain only the portion after "@" as the username. Then, "text" is dropped from the input, and the
newusername is "abc.com". To apply username truncation function to a named-attribute format, see Configuring
AAA Attribute Format Function, on page 42.
Release 4.3.x
32 OL-28375-03
Creating Attributes of Specific Format
BNG supports the use of configurable AAA attributes. The configurable AAA attributes have specific
user-defined formats. The following sections list some of the configurable AAA attributes used by BNG.
Username
BNG has the ability to construct AAA username and other format-supported attributes for subscribers using
MAC address, circuit-ID, remote-ID, and DHCP Option-60 (and a larger set of values available in CLI). The
DHCP option-60 is one of the newer options that is communicated by the DHCP client to the DHCP server
in its requests; it carries Vendor Class Identifier (VCI) of the DHCP client's hardware.
The MAC address attribute is specified in the CLI format in either of these forms:
mac-address: for example, 0000.4096.3e4a
mac-address-ietf: for example, 00-00-40-96-3E-4A
mac-address-raw: for example, 000040963e4a
An example of constructing a username in the form "mac-address@vendor-class-ID" is:
aaa attribute format USERNAME-FORMAT format-string %s@%s mac-address dhcp-vendor-class
NAS-Port-ID
The NAS-Port-IDis constructed by combining BNGport information and access-node information. The BNG
port information consists of a string in this form:
"eth phy_slot/phy_subslot/phy_port:XPI.XCI"
For 802.1Q tunneling (QinQ), XPI is the outer VLAN tag and XCI is the inner VLAN tag.
If the interface is QinQ, the default format of nas-port-ID includes both the VLAN tags; if the interface is
single tag, it includes a single VLAN tag.
In the case of a single VLAN, only the outer VLAN is configured, using this syntax:
<slot>/<subslot>/<port>/<outer_vlan>
In the case of QinQ, the VLAN is configured using this syntax:
<slot>/<subslot>/<port>/<inner_vlan>.<outer_vlan>
The nas-port-IDcommand is extended to use the 'nas-port-type' option so that the customized format (configured
with the command shown above) can be used on a specific interface type (nas-port-type). The extended
nas-port-ID command is:
aaa radius attribute nas-port-id format <FORMAT_NAME> [type <NAS_PORT_TYPE>]
If 'type' option is not specified, then the nas-port-ID for all interface types is constructed according to the
format name specified in the command. An example of constructing a maximum 128 byte NAS-Port-ID, by
combining the BNG port information and Circuit-ID is:
aaa attribute format NAS-PORT-ID-FORMAT1 format-string length 128 eth %s/%s/%s:%s.%s %s
physical-slot physical-subslot physical-port outer-vlan-Id inner-vlan-id circuit-id-tag
An example of constructing the NAS-Port-ID from just the BNG port information, and with "0/0/0/0/0/0"
appended at the end for circuit-ID, is:
aaa attribute format NAS-PORT-ID-FORMAT2 format-string eth %s/%s/%s:%s.%s 0/0/0/0/0/0
physical-slot physical-subslot physical-port outer-vlan-Id inner-vlan-id
4.3.x
OL-28375-03 33
An example of constructing the NAS-Port-ID from just the Circuit-ID is:
aaa attribute format NAS-PORT-ID-FORMAT3 format-string %s circuit-id-tag
The NAS-Port-ID formats configured in the above examples, can be specified in the nas-port-ID command,
thus:
For IPoEoQINQ interface:-
aaa radius attribute nas-port-id format NAS-PORT-ID-FORMAT1 type 41
For Virtual IPoEoQINQ interface:-
For IPOEoE interface:-
NAS-Port-Type on Interface or VLAN Sub-interface
In order to have different production models for subscribers on the same BNG router, but different physical
interfaces of same type, the NAS-Port-Type is made configurable for each physical interface, or VLAN
sub-interface. With a different NAS-Port-Type value configured on the interface, the NAS-Port and
NAS-Port-IDgets formatted according to the formats defined globally for the newNAS-Port-Type configured
on the interface, instead of the actual value of NAS-Port-Type that the interface has. This in turn sends different
formats of NAS-Port, NAS-Port-ID and NAS-Port-Type to the RADIUS server for the subscribers under
different production models.
In the case of sub-interfaces, the hierarchy to be followed in deciding the format of NAS-Port-Type to be sent
to the RADIUS server is:
1 Verify whether the NAS-Port-Type is configured on the sub-interface in which the subscriber session
arrives.
2 If NAS-Port-Type is not configured on the sub-interface, verify whether it is configured on the main
physical interface.
The format of NAS-Port or NAS-Port-ID is based on the NAS-Port-Type retrieved in Step 1 or Step 2.
3 If NAS-Port-Type is configured on neither the sub-interface nor the main physical interface, the format
of NAS-Port or NAS-Port-ID is based on the format of the default NAS-Port-Type of the sub-interface.
4 If a NAS-Port or NAS-Port-ID format is not configured for the NAS-Port-Type retrieved in steps 1, 2 or
3, the format of NAS-Port or NAS-Port-ID is based on the default formats of NAS-Port or NAS-Port-ID.
Use this command to configure NAS-Port-Type per interface or VLAN sub-interface:
aaa radius attribute nas-port-type <nas-port-type>
where:
<nas-port-type> is either a number ranging from 0 to 44, or a string specifying the nas-port-type.
Refer Configuring RADIUS Attribute Nas-port-type, on page 40.
Calling-Station-ID and Called-Station-ID
BNG supports the use of configurable calling-station-ID and called-station-ID. The calling-station-ID is a
RADIUS attribute that uses Automatic Number Identification (ANI), or similar technology. It allows the
network access server (NAS) to send to the Access-Request packet, the phone number from which the call
came from. The called-station-ID is a RADIUS attribute that uses Dialed Number Identification (DNIS), or
similar technology. It allows the NAS to send to the Access-Request packet, the phone number that the user
called from.
Release 4.3.x
34 OL-28375-03
The command used to configure the calling-station-ID and called-station-ID attributes is:
aaa radius attribute calling-station-id format <FORMAT_NAME>
aaa radius attribute called-station-id format <FORMAT_NAME>
Examples of constructing calling-station-ID from mac-address, remote-ID, and circuit-ID are:
aaa radius attribute calling-station-id format CLID-FORMAT
aaa attribute format CLID-FORMAT format-string %s:%s:%s client-mac-address-ietf remote-id-tag
circuit-id-tag
Examples of constructing called-station-ID from mac-address, remote-ID, and circuit-ID are:
aaa radius attribute called-station-id format CLDID-FORMAT
aaa attribute format CLDID-FORMAT format-string %s:%s client-mac-address-raw circuit-id-tag
NAS-Port Format
NAS-Port is a 4-byte value that has the physical port information of the Broadband Remote Access Server
(BRAS), which connects the Access Aggregation network to BNG. It is used both by Access-Request packets
and Accounting-Request packets. To uniquely identify a physical port on BRAS, multiple pieces of information
such as shelf, slot, adapter, and so on is used along with the port number. Aconfigurable format called format-e
is defined to allowindividual bits or group of bits in 32 bits of NAS-Port to represent or encode various pieces
that constitute port information.
Individual bits in NAS-Port can be encoded with these characters:
Zero: 0
One: 1
PPPoX slot: S
PPPoX adapter: A
PPPoX port: P
PPPoX VLAN Id: V
PPPoX VPI: I
PPPoX VCI: C
Session-Id: U
PPPoX Inner VLAN ID: Q
aaa radius attribute nas-port format e [string] [type {nas-port-type}]
The above command is used to configure a format-e encode string for a particular interface of NAS-Port type
(RADIUS attribute 61). The permissible nas-port type values are:
Whether value can be
configured on the
interface configuration
mode
derived from
associated interface
Values Nas-port-types
Yes No 0 ASYNC
4.3.x
OL-28375-03 35
configured on the
interface configuration
mode
derived from
associated interface
Values Nas-port-types
Yes No 1 SYNC
Yes No 2 ISDN
Yes No 3 ISDN_V120
Yes No 4 ISDN_V110
Yes No 5 VIRTUAL
Yes No 6 ISDN_PIAFS
Yes No 9 X75
Yes No 15 ETHERNET
Yes No 30 PPPATM
Yes No 31 PPPOEOA
Yes Yes 32 PPPOEOE
Yes Yes 33 PPPOEOVLAN
Yes Yes 34 PPPOEOQINQ
Yes Yes 35 VIRTUAL_PPPOEOE
Yes Yes 36 VIRTUAL_PPPOEOVLAN
Yes Yes 37 VIRTUAL_PPPOEOQINQ
Yes No 38 IPSEC
Yes Yes 39 IPOEOE
Yes Yes 40 IPOEOVLAN
Yes Yes 41 IPOEOQINQ
Yes Yes 42 VIRTUAL_IPOEOE
Yes Yes 43 VIRTUAL_IPOEOVLAN
Yes Yes 44 VIRTUAL_IPOEOQINQ
Release 4.3.x
36 OL-28375-03
Examples:
For non-bundle: GigabitEthernet0/1/2/3.11.pppoe5
where:
PPPoEoQinQ (assuming 2 vlan tags): interface-type
1: slot
2: adapter
3: port
vlan-ids: whatever the outer and inner vlan-ids received in the PADR were
5: session-id
aaa radius attribute nas-port format e SSAAPPPPQQQQQQQQQQVVVVVVVVVVUUUU type 34
Generated NAS-Port: 01100011QQQQQQQQQQVVVVVVVVVV0101
For bundle: Bundle-Ether17.23.pppoe8
where:
Virtual-PPPoEoQinQ (assuming 2 vlan tags): interface-type
0: slot
0: adapter
17 (bundle-id): port
Vlan-Ids: whatever the outer and inner vlan-ids received in the PADR were.
8: session-id
aaa radius attribute nas-port format e PPPPPPQQQQQQQQQQVVVVVVVVVVUUUUUU type 37
Generated NAS-Port: 010001QQQQQQQQQQVVVVVVVVVV000101
NAS-port format for IP/DHCP sessions are represented in these examples:
For IPoEoVLAN interface type:
aaa radius attribute nas-port format e SSAAAPPPPPVVVVVVVVVVVVVVVVVVVVVV type 40
For IPoEoQinQ:
aaa radius attribute nas-port format e SSAAAPPPPPQQQQQQQQQQQVVVVVVVVVVV type 41
For virtual IPoEoVLAN:
aaa radius attribute nas-port format e PPPPPPPPVVVVVVVVVVVVVVVVUUUUUUUU type 43
NAS-port format for PPPoE sessions are represented in these examples:
For PPPoEoVLAN interface type:
aaa radius attribute nas-port format e SSAAAPPPPPVVVVVVVVVVVVVVVVVVUUUU type 33
For Virtual PPPoEoVLAN:.
aaa radius attribute nas-port format e PPPPPPPPVVVVVVVVVVVVVVVVUUUUUUUU type 36
If a NAS-Port format is not configured for a NAS-Port-Type, the system looks for a default CLI
configuration for the NAS-Port format. In the absence of both these configurations, for sessions with that
particular NAS-Port-Type, the NAS-Port attribute is not sent to the RADIUS server.
Note
Configuring RADIUS Attribute List
Perform this task to create a RADIUS attribute list that is used for filtering authorization and accounting
attributes.
4.3.x
OL-28375-03 37
SUMMARY STEPS
1. configure
2. radius-server attribute list listname
3. attribute list_of_radius_attributes
4. attribute vendor-id vendor-type number
5. vendor-type vendor-type-value
DETAILED STEPS
Example:
Step 1
Defines the name of the attribute list. radius-server attribute list listname
Example:
RP/0/RSP0/CPU0:router(config)# radius-server attribute
list l1
Step 2
Populates the list with radius attributes. attribute list_of_radius_attributes Step 3
Example:
RP/0/RSP0/CPU0:router(config-attribute-filter)# attribute
a1, a2
For more information about supported
attributes, see RADIUS Attributes, on page
291 .
Note
Configures the attribute filtering to be applied to
vendor specific attributes (VSAs) by allowing vendor
attribute vendor-id vendor-type number
Example:
RP/0/RSP0/CPU0:router(config)# attribute vendor-id 6456
Step 4
specific information for VSAs to be specified in
radius attribute list CLI. Vendor specific information
comprises of vendor-id, vendor-type, and optional
attribute name in case of Cisco generic VSA. The
vendor-id ranges from 0 to 4294967295.
Configures the vendor specific information such as
the vendor-type to be specified in radius attribute list.
The range of the vendor-type value is from 1 to 254.
vendor-type vendor-type-value
Example:
RP/0/RSP0/CPU0:router(config-attribute-filter-vsa)#
vendor-type 54
Step 5
Release 4.3.x
38 OL-28375-03
Configuring RADIUS Attribute List: An example
configure
radius-server attribute list list_! attribute B C
attribute vendor-id vendor-type 10
vendor-type 30
!
end
Configuring RADIUS Attribute Format
Perform this task to the define RADIUS attribute format for the nas-port attribute, and apply a predefined
format on nas-port-ID attribute.
SUMMARY STEPS
1. configure
2. aaa radius attribute
3. nas-port format e string type nas-port-type value
4. nas-port-id format format name
DETAILED STEPS
Example:
Step 1
4.3.x
OL-28375-03 39
Configuring RADIUS Attribute Format
Configures the AAA radius attribute. aaa radius attribute
Example:
RP/0/RSP0/CPU0:router(config)# aaa radius
attribute
Step 2
Configures the format for nas-port attribute. The string represents
a 32 character string representing the format to be used. The
nas-port-value ranges from 0 to 44.
nas-port format e string type nas-port-type value
Example:
RP/0/RSP0/CPU0:router(config)# nas-port
format e format1 type 30
Step 3
Applies a predefined format to the nas-port-ID attribute. nas-port-id format format name
Example:
RP/0/RSP0/CPU0:router(config)# nas-port-id
format format2
Step 4
session.
Configuring RADIUS Attribute Format: An example
configure
aaa radius attribute
nas-port format e abcd type 40
nas-port-id format ADEF
!
end
Configuring RADIUS Attribute Nas-port-type
Performthis task to configure RADIUS Attribute nas-port-type on a physical interface or VLANsub-interface:
Release 4.3.x
40 OL-28375-03
SUMMARY STEPS
1. configure
2. interface type interface-name
3. aaa radius attribute nas-port-type {value | name}
DETAILED STEPS
Example:
Step 1
Enters the interface configuration mode. interface type interface-name
Example:
RP/0/RSP0/CPU0:router(config)# interface
gigabitEthernet 0/0/0/0
Step 2
Configures the RADIUS Attribute nas-port-type value. aaa radius attribute nas-port-type {value |
name}
Step 3
The range of value is from 0 to 44.
Example:
RP/0/RSP0/CPU0:router(config-if)# aaa
radius attribute nas-port-type 30
See table in NAS-Port Format, on page 35, for permissible
nas-port-type values within this range.
or
RP/0/RSP0/CPU0:router(config-if)# aaa
radius attribute nas-port-type Ethernet
session.
Configuring RADIUS Attribute Nas-port-type: An example
configure
interface gigabitEthernet 0/0/0/0
aaa radius attribute nas-port-type Ethernet
4.3.x
OL-28375-03 41
!
end
Configuring AAA Attribute Format Function
Perform this task to configure a function for the AAA attribute format. The function is for stripping the
user-name till the delimiter.
SUMMARY STEPS
1. configure
2. aaa attribute format format-name
3. username-strip prefix-delimiter prefilx_delimiter
DETAILED STEPS
Example:
Step 1
Specifies the format name for which the function is defined. aaa attribute format format-name
Example:
attribute format red
Step 2
Configures the function to strip the username preceding the prefix
delimiter, which is @.
username-stripprefix-delimiter prefilx_delimiter
Example:
RP/0/RSP0/CPU0:router(config-id-format)#
username-strip prefix-delimiter @
Step 3
session.
Release 4.3.x
42 OL-28375-03
Configuring AAA Attribute Format Function
Configuring AAA Attribute Format Function: An example
configure
aaa attribute format red
username-strip prefix-delimiter @
!
!
end
Making RADIUS Server Settings
In order to make BNG interact with the RADIUS server, certain server specific settings must be made on the
BNG router. This table lists some of the key settings:
Description Settings
Defines the RADIUS server details to which BNG will connect. Server host
Defines which attribute list is to be used. Attribute list
Defines the encryption status. Server key
Defines the criteria that is used to mark a RADIUS server as dead. Dead criteria
Defines the number of retries the BNG makes to send data to RADIUS server. Retransmit value
Defines how long BNG waits for the RADIUS server to reply. Timeout value
Defines the duration after which automated testing will start and the username to be
tested.
Automated
testing
Allows RADIUS packets to be marked with a specific Differentiated Services Code Point
(DSCP) value.
IP DSCP
For more making RADIUS server settings, see Configuring RADIUS Server Settings, on page 43.
For more making specific automated testing settings, see Configuring Automated Testing, on page 47.
For more making specific IP DSCP settings, see Setting IP DSCP for RADIUS Server, on page 49.
Restriction
The service profile push or asynchronously pushing a profile to the system is not supported. To download a
profile fromRadius, the profile must be requested initially as part of the subscriber request. Only service-update
is supported and can be used to change a service that was previously downloaded.
Configuring RADIUS Server Settings
Perform this task to make RADIUS server specific settings on the BNG router.
4.3.x
OL-28375-03 43
Making RADIUS Server Settings
SUMMARY STEPS
1. configure
2. radius-server host host_name acct-port accounting_port_number auth-port authentication_port_number
3. radius-server attribute list list_name attribute_list
4. radius-server key 7 encrypted_text
5. radius-server disallow null-username
6. radius-server dead-criteria time value
7. radius-server dead-criteria tries value
8. radius-server deadtime limit
9. radius-server ipv4 dscp codepoint_value
10. radius-server load-balance method least-outstanding ignore-preferred-server batch-size size
11. radius-server retransmit retransmit_value
12. radius-server source-port extended
13. radius-server timeout value
14. radius-server vsa attribute ignore unknown
15. radius source-interface Loopback value vrf vrf_name
DETAILED STEPS
Example:
Step 1
Specifies the radius server and its host name.
Configures the UDP port for RADIUS accounting
radius-server host host_name acct-port accounting_port_number
auth-port authentication_port_number
Step 2
and authentication requests. The accounting and
Example:
RP/0/RSP0/CPU0:router(config)# radius-server host 1.2.3.4
acct-port 455 auth-port 567
authentication port numbers range from 0 to 65535.
If no value is specified, then the default is 1645 for
the auth-port and 1646 for the acct-port.
Specifies the radius server attributes list, and
customizes the selected radius attributes.
radius-server attribute list list_name attribute_list
Example:
RP/0/RSP0/CPU0:router(config)# radius-server attribute
list rad_list a b
Step 3
Specifies the per-server encryption key that overrides
the default, and takes the value 0 or 7, which indicates
that the unencrypted key will follow.
radius-server key 7 encrypted_text
Example:
RP/0/RSP0/CPU0:router(config-radius-host)# radius-server
key 7 rngiry
Step 4
Release 4.3.x
44 OL-28375-03
Specifies that the null-username is disallowed for the
radius server.
radius-server disallow null-username
Example:
RP/0/RSP0/CPU0:router(config)# radius-server disallow
null-username
Step 5
Specifies the dead server detection criteria for a
configured RADIUS server. The time (in seconds)
radius-server dead-criteria time value
Example:
RP/0/RSP0/CPU0:router(config)# radius-server
dead-criteria time 40
Step 6
specifies the minimum time that must elapse since a
response is received from this RADIUS server.
Specify the value for the number of consecutive
timeouts that must occur on the router before the
radius-server dead-criteria tries value
Example:
dead-criteria tries 50
Step 7
RADIUS server is marked as dead. The value ranges
from 1 to 100.
Specifies the time in minutes for which a RADIUS
server is marked dead. The deadtime limit is specified
radius-server deadtime limit
Example:
RP/0/RSP0/CPU0:router(config)# radius-server deadtime 67
Step 8
in minutes and ranges from 1 to 1440. If no value is
specified, the default is 0.
Allows radius packets to be marked with a specific
differentiated services code point (DSCP) value. This
code point value ranges from 0 to 63.
radius-server ipv4 dscp codepoint_value
Example:
RP/0/RSP0/CPU0:router(config)# radius-server ipv4 dscp
45
Step 9
Configures the radius load-balancing options by
picking the server with the least outstanding
radius-server load-balance method least-outstanding
ignore-preferred-server batch-size size
Step 10
transactions. This load-balancing method uses the
Example:
RP/0/RSP0/CPU0:router(config)# radius-server load-balance
batch-size for the selection of the server. The size
ranges from 1 to 1500. If no value is specified, the
default is 25.
method least-outstanding ignore-preferred-server
batch-size 500
Specifies the number of retries to the active server.
The retransmit value indicates the number of retries
radius-server retransmit retransmit_value
Example:
RP/0/RSP0/CPU0:router(config)# radius-server retransmit
45
Step 11
in numeric and ranges from 1 to 100. If no value is
specified, then the default is 3.
Configures BNG to use a total of 200 ports as the
source ports for sending out RADIUS requests.
radius-server source-port extended
Example:
RP/0/RSP0/CPU0:router(config)# radius-server source-port
extended
Step 12
4.3.x
OL-28375-03 45
Specifies the time to wait for a radius server to reply.
The value is in seconds and ranges from 1 to 1000.
The default is 5.
radius-server timeout value
Example:
RP/0/RSP0/CPU0:router(config)# radius-server timeout
Step 13
Ignores the unknown vendor-specific attributes for
the radius server.
radius-server vsa attribute ignore unknown
Example:
RP/0/RSP0/CPU0:router(config)# radius-server vsa
attribute ignore unknown
Step 14
Specifies loopback interface for source address in
RADIUS packets. The value ranges from0 to 65535.
radius source-interface Loopback value vrf vrf_name
Example:
RP/0/RSP0/CPU0:router(config)# radius source-interface
Loopback 655 vrf vrf_1
Step 15
Configuring RADIUS Server Settings: Examples
\\Configuring RADIUS Server Options
configure
radius-server attribute list list1 a b
radius-server dead-criteria time 100
radius-server deadtime 30
radius-server host 1.2.3.4 acct-port 655 auth-port 566
radius-server ipv4 dscp 34
radius-server key 7 ERITY$
radius-server load-balance method least-outstanding ignore-preferred-server batch-size 25
radius-server retransmit 50
radius-server source-port extended
radius-server timeout 500
!
!
end
\\Configuring RADIUS Attribute List
radius-server attribute list list_! attribute B C
attribute vendor-id vendor-type 10
Release 4.3.x
46 OL-28375-03
vendor-type 30
!
end
\\Configuring RADIUS Server Host
configure
radius-server host 1.3.5.7 acct-port 56 auth-port 66
idle-time 45
ignore-acct-port
ignore-auth-port 3.4.5.6
key 7 ERWQ
retransmit 50
test username username
timeout 500
!
end
\\Configuring RADIUS Server Key
configure
radius-server key 7 ERWQ
!
end
\\Configuring Load Balancing for RADIUS Server
configure
radius-server load-balance method least-outstanding batch-size 25
!
end
\\Ignoring Unknown VSA Attributes in RADIUS Server
configure
!
end
\\Configuring Dead Criteria for RADIUS Server
configure
radius-server dead-criteria time 60
radius-server dead-criteria tries 60
!
end
\\Configuring Disallow Username
configure
!
end
\\Setting IP DSCP for RADIUS Server
configure
radius-server ipv4 dscp default
!
end
Configuring Automated Testing
Perform this task to test if the external RADIUS server is UP or not.
4.3.x
OL-28375-03 47
SUMMARY STEPS
1. configure
2. radius-server idle-time idle_time
3. radius-server test username username
DETAILED STEPS
Example:
Step 1
Specifies the idle-time after which the automated test
should start. The idle time is specified in minutes, and
ranges from 1 to 60.
radius-server idle-time idle_time
Example:
idle-time 45
Step 2
Specifies the username to be tested for the automated
testing functionality.
radius-server test username username
Example:
test username user1
Step 3
Configuring Automated Testing: An example
configure
radius-server idle-time 60
radius-server test username user_1
!
end
Release 4.3.x
48 OL-28375-03
Setting IP DSCP for RADIUS Server
Perform this task to set IP differentiated services code point (DSCP) for RADIUS server.
SUMMARY STEPS
1. configure
2. radius-server ipv4 dscp codepoint_value
3. radius-server ipv4 dscp default
DETAILED STEPS
Example:
Step 1
Allows radius packets to be marked with a specific differentiated
services code point (DSCP) value that replaces the outdated IP
radius-server ipv4 dscp codepoint_value
Example:
ipv4 dscp 45
Step 2
precedence, a 3-bit field in the Type of Service byte of the IP header
originally used to classify and prioritize types of traffic. This code
point value ranges from 0 to 63.
Matches the packets with default dscp (000000). radius-server ipv4 dscp default
Example:
ipv4 dscp default
Step 3
Yes Saves configuration changes and exits the
4.3.x
OL-28375-03 49
Setting IP DSCP for RADIUS Server
Setting IP DSCP for RADIUS Server: An example
configure
radius-server ipv4 dscp default
!
end
Balancing Transaction Load on the RADIUS Server
The RADIUS load-balancing feature is a mechanism to share the load of RADIUS access and accounting
transactions, across a set of RADIUS servers. Each AAA request processing is considered to be a transaction.
BNG distributes batches of transactions to servers within a server group.
When the first transaction for a new is received, BNG determines the server with the lowest number of
outstanding transactions in its queue. This server is assigned that batch of transactions. BNG keeps repeating
this determination process to ensure that the server with the least-outstanding transactions always gets a new
batch. This method is known as the least-outstanding method of load balancing.
You can configure the load balancing feature either globally, or for RADIUS servers that are part of a server
group. In the server group, if a preferred server is defined, you need to include the keyword
"ignore-preferred-server" in the load-balancing configuration, to disable the preference.
For configuring the load balancing feature globally, see Configuring Load Balancing for Global RADIUS
Server Group, on page 50.
For configuring the load balancing feature on RADIUS servers that are part of a named server group, see
Configuring Load Balancing for a Named RADIUS Server Group, on page 51.
Configuring Load Balancing for Global RADIUS Server Group
Perform this task to activate the load balancing function for the global RADIUS server group. As an example,
in this configuration the preferred server is set to be ignored.
SUMMARY STEPS
1. configure
2. radius-server load-balance method least-outstanding batch-size size
3. radius-server load-balance method least-outstanding ignore-preferred-server batch-size size
DETAILED STEPS
Example:
Step 1
Release 4.3.x
50 OL-28375-03
Balancing Transaction Load on the RADIUS Server
Configures the radius load-balancing options by picking the server
with the least-outstanding transactions. This load-balancing method
radius-server load-balance method
least-outstanding batch-size size
Step 2
uses the batch-size for the selection of the server. The size ranges
from 1 to 1500. If no value is specified, the default is 25.
Example:
load-balance method least-outstanding
batch-size 500
Configures the radius load-balancing options by disabling the
preferred server for this Server Group. This load-balancing method
radius-server load-balance method
least-outstanding ignore-preferred-server
batch-size size
Step 3
Example:
ignore-preferred-server batch-size 500
session.
Configuring Load Balancing for RADIUS Server: An example
configure
radius-server load-balance method least-outstanding batch-size 25
!
end
Configuring Load Balancing for a Named RADIUS Server Group
Perform this task to activate the load balancing function for a named RADIUS server group. As an example,
in this configuration the preferred server is set to be ignored.
4.3.x
OL-28375-03 51
SUMMARY STEPS
1. configure
2. aaa group server radius server_group_name load-balance method least-outstanding batch-size size
3. aaa group server radius server_group_name load-balance method least-outstanding
DETAILED STEPS
Example:
Step 1
Configures the radius load-balancing options by picking the server
with the least-outstanding transactions. This load-balancing method
aaa group server radius server_group_name
load-balance method least-outstanding batch-size
size
Step 2
Example:
RP/0/RSP0/CPU0:router(config)# aaa group
server radius sg1 load-balance method
least-outstanding batch-size 500
Configures the radius load-balancing options by disabling the
preferred server for this Server Group. This load-balancing method
aaa group server radius server_group_name
Step 3
Example:
RP/0/RSP0/CPU0:router(config)# aaa group
server radius sg1 load-balance method
least-outstanding ignore-preferred-server
batch-size 500
session.
Release 4.3.x
52 OL-28375-03
Throttling of RADIUS Records
The Throttling of AAA (RADIUS) records is a mechanism to avoid RADIUS congestion and instability. This
function is useful in situations when there is insufficient bandwidth to accommodate a sudden burst of AAA
requests generated by the BNG for the RADIUS server.
While configuring throttling, a threshold rate, which corresponds to the maximum number of outstanding
requests, is defined. It is possible to configure independent throttling rates for access (authentication and
authorization) and accounting requests. After a threshold value is reached for a server, no further requests of
that type are sent to the server. However, for the pending requests, a retransmit timer is started, and if the
outstanding request count (which is checked after every timer expiry), is less than the threshold, then the
request is sent out.
As a session may timeout due to throttle on the access requests, a limit is set for the number of retransmit
attempts. After this limit is reached, further access requests are dropped. Throttled accounting requests,
however, are processed through the server-group failover process.
The throttling feature can be configured globally, or for a server-group. However, the general rule of
configuration preference is that the server-group configuration overrides global configuration, if any.
The syntax for the throttling CLI command is:
radius-server throttle {[accounting THRESHOLD] [access THRESHOLD [access-timeout
NUMBER_OF-TIMEOUTS]]}
where:
accounting THRESHOLDSpecifies the threshold for accounting requests. The range is from 0 to
65536. The default is 0, and indicates that throttling is disabled for accounting requests.
access THRESHOLDSpecifies the threshold for access requests. The range is from 0 to 65536. The
default is 0, and indicates that throttling is disabled for accounting requests.
access-timeout NUMBER_OF-TIMEOUTSSpecifies the number of consecutive timeouts that must
occur on the router, after which access-requests are dropped. The range of is from 0 to 10. The default
is 3.
By default, the throttling feature is disabled on BNG. Note
For activating throttling globally, see Configuring RADIUS Throttling Globally, on page 53.
For activating throttling on a server group, see Configuring RADIUS Throttling on a Server Group, on page
55.
Configuring RADIUS Throttling Globally
Perform this task to activate RADIUS throttling globally.
4.3.x
OL-28375-03 53
Throttling of RADIUS Records
SUMMARY STEPS
1. configure
2. radius-server throttle access threshold_value
3. radius-server throttle access threshold_value access-timeout value
4. radius-server throttle access threshold_value access-timeout value accounting threshold_value
5. radius-server throttle accounting threshold_value access value access-timeout value
DETAILED STEPS
Example:
Step 1
Controls the number of access requests sent to a RADIUS server.
The threshold value denotes the number of outstanding access
radius-server throttle access threshold_value
Example:
throttle access 10
Step 2
requests after which throttling should be performed. The range is
from 0 to 65535, and the preferred value is 100.
Specifies the number of timeouts, after which a throttled access
request is dropped. The value denotes the number of timeouts for
a transaction. The range is from 1 to 10, and the default is 3.
access-timeout value
Example:
throttle access 10 access-timeout 5
Step 3
Controls the number of access timeout requests sent to a RADIUS
server. The threshold value denotes the number of outstanding
access-timeout value accounting threshold_value
Step 4
accounting transactions after which throttling should be performed.
The range is from 0 to 65535, and the preferred value is 100.
Example:
throttle access 10 access-timeout 5
accounting 10
Controls the number of accounting requests sent to a RADIUS
server. The threshold value denotes the number of outstanding
radius-server throttle accounting threshold_value
access value access-timeout value
Step 5
accounting transactions after which throttling should be performed.
The value ranges between 0 to 65535 and the preferred value is 100.
Example:
throttle accounting 56 access 10
access-timeout 5
Release 4.3.x
54 OL-28375-03
Configuring RADIUS Throttling Globally
session.
Configuring RADIUS Throttling Globally: An example
configure
radius-server throttle access 10 access-timeout 5 accounting 10
!
end
Configuring RADIUS Throttling on a Server Group
Perform this task to activate RADIUS throttling on a server group.
SUMMARY STEPS
1. configure
2. aaa group server radius server_group_name
3. server hostname acct-port acct_port_value auth-port auth_port_value
4. throttle access threshold_value access-timeout value accounting threshold_value
DETAILED STEPS
Example:
Step 1
Configures the AAA(RADIUS) server-group definition. aaa group server radius server_group_name
Example:
RP/0/RSP0/CPU0:router(config)# aaa group server radius
SG1
Step 2
4.3.x
OL-28375-03 55
Configuring RADIUS Throttling on a Server Group
Configures a RADIUS server accounting or
authentication port with either the IP address or hostname
server hostname acct-port acct_port_value auth-port
auth_port_value
Step 3
(as specified). The accounting port number and the
authentication port number ranges from 0 to 65535.
Example:
RP/0/RSP0/CPU0:router(config-sg-radius)# server
99.1.1.10 auth-port 1812 acct-port 1813
Configures the RADIUS throttling options to control the
number of access and accounting requests sent to a
throttle access threshold_value access-timeout value accounting
threshold_value
Step 4
RADIUS server. The threshold value denotes the number
Example:
RP/0/RSP0/CPU0:router(config-sg-radius)# radius-server
throttle access 10 access-timeout 5 accounting 10
of outstanding access requests or accounting transactions
after which throttling should be performed. The range is
from 0 to 65535, and for both access and accounting
requests the preferred value is 100.
commitSaves the configuration changes, and remains
within the configuration session.
Configuring RADIUS Throttling on a Server Group: An example
configure
aaa group server radius SG1
server 99.1.1.10 auth-port 1812 acct-port 1813
radius-server throttle access 10 access-timeout 5 accounting 10
!
end
RADIUS Change of Authorization (CoA) Overview
The Change of Authorization (CoA) function allows the RADIUS server to change the authorization settings
for a subscriber who is already authorized. CoA is an extension to the RADIUS standard that allows sending
asynchronous messages from RADIUS servers to a RADIUS client, like BNG.
A CoA server can be a different from the RADIUS server. Note
Release 4.3.x
56 OL-28375-03
To identify the subscriber whose configuration needs to be changed, a RADIUS CoA server supports and
uses a variety of keys (RADIUS attributes) such as Accounting-Session-ID, Username, IP-Address, and
ipv4:vrf-id.
The RADIUS CoA supports:
account-logon When a user logs into a network, an external web portal that supports CoA sends an
account-logon request to BNG with the user's credentials (username and password). Account-logon on
BNG then attempts to authenticate the user through RADIUS with those credentials.
account-logoffBNGprocesses the account-logoff request as a disconnect event for the subscriber and
terminates the session.
The RADIUS CoA server does not differentiate between originators of the disconnect
event. Hence, when the BNG receives an account-logoff request from the RADIUS
CoA server, for both a user-initiated and an administrator-initiated request, the
Acct-Terminate-Cause to be sent to the RADIUS server is always set as Admin-Reset.
Note
account-update BNG parses and applies the attributes received as part of the CoA profile. Only
subscriber-specific attributes are supported and applied on the user profile.
activate-service BNG starts a predefined service on a subscriber. The service settings can either be
defined locally by a dynamic template, or downloaded from the RADIUS server.
deactivate-service BNG stops a previously started service on the subscriber, which is equivalent to
deactivating a dynamic-template.
For a list of supported Vendor-Specific Attributes for account operations, see Vendor-Specific Attributes for
Account Operations, on page 298.
Service Activate from CoA
BNG supports activating services through CoA requests. The CoA service-activate command is used for
activating services. The CoA request for the service activate should contain these attributes:
"subscriber:command=activate-service" Cisco VSA
"subscriber:service-name=<service name>" Cisco VSA
Other attributes that are part of the service profile
The "<subscriber:sa=<service-name>" can also be used to activate services from CoA and through RADIUS.
Duplicate service activate requests can be sent to BNG from the CoA server. BNG does not take any action
on services that are already activated. BNGsends a CoAACKmessage to the CoAserver under these scenarios:
When a duplicate request with identical parameters comes from the CoA for a service that is already
active.
When a duplicate request with identical parameters comes fromthe CoAto apply a parameterized service.
BNG sends a CoA NACK message to the CoA server with an error code as an invalid attribute under these
scenarios:
4.3.x
OL-28375-03 57
When a request comes from the CoA to deactivate a non-parameterized service that is not applied to the
session.
When a request comes from the CoA to deactivate a parameterized service that is not applied to the
session.
When a duplicate request to apply a parameterized service is made with non-identical parameters from
the CoA.
When a request with non-identical parameters comes from CoA to deactivate a parameterized service.
Service Update from CoA
The service update feature allows an existing service-profile to be updated with a new RADIUS attribute list
representing the updated service. This impacts any subscriber who is already activated with the service and
new subscriber who activate the service in the future. The new CoA service-update command is used for
activating this feature. The CoA request for the service update should have these attributes:
"subscriber:command=service-update" Cisco VSA
"subscriber:service-name=<service name>" Cisco VSA
Other attributes that are part of the service profile
A service update CoA should have a minimum of these attributes:
vsa cisco generic 1 string "subscriber:command=service-update"
vsa cisco generic 1 string "subscriber:service-name=<service name>"
Web Logon with RADIUS Based CoA
To support Web Logon, a set of Policy Rule Events need to be configured in an ordered manner. These events
are as follows:
session-start:
On the start of a session, a subscriber is setup to get internet connectivity. The service is activated
to redirect HTTP traffic to a Web portal for web-based logon.
Start the timer with duration for the maximum waiting period for authentication.
account-logon The Web portal collects the user credentials such as username and password and
triggers a CoAaccount-logon command. When this event is triggered, subscriber username and password
are authenticated by the RADIUS server. Once the authentication is successful, the HTTP redirect service
is deactivated, granting user access to already connected internet setup. Also, the timer established in
session-start must be stopped. However, if the authentication fails during account-logon, BNG sends a
NAK CoA request, allowing for further authentication attempts to take place.
timer expiry When the timer expires, the subscriber session is disconnected based on the configuration.
Service Accounting
Accounting records for each service enabled on a subscriber can be sent to the configured RADIUS server.
These records can include service-start, service-stop, and service-interim records containing the current state
Release 4.3.x
58 OL-28375-03
Service Accounting
of the service and any associated counters. This feature is the Service Accounting feature. Service accounting
records are consolidated accounting records that represent the collection of features that make up a service as
part of a subscriber session.
Service accounting starts when a subscriber session comes up with a service enabled on it. This can happen
through a dynamic template applied through a control policy, through access-accept (AA) messages when
the session is authorized, or through a change of authorization (CoA), when a new service is applied on a
subscriber session. Service accounting stops either when the session is terminated, or a service is removed
fromthe session through CoA, or some other event that deactivates the service. Start records have no counters;
interimand stop records with QoS counters are generated when service accounting is enabled for QoS. Interim
accounting records can be generated, in between start and stop accounting, as an option with a pre-defined
periodic interval. When the interimperiod is zero, interimaccounting records are not created. Different interim
intervals are based on every service for each session. Service accounting is enabled on each template, based
on the configuration.
The policy-map associated to a dynamic template can be edited to change the service parameters. However,
this does not update the accounting records. Therefore, to generate all the accounting records accurately,
it is recommended that a new service with all the required service parameters be created and associated
to the new service, through a CoA.
Note
For service accounting, statistics for ingress and egress QoS policies, which are applied under each service
for a given subscriber, may need to be reported as part of the accounting interim and stop records. For each
service, these QoS counters can be reported as part of the accounting records:
BytesIn Aggregate of bytes matching all classes of the ingress QoS policy for the service minus the
policer drops.
PacketsIn Aggregate of packets matching all classes of the ingress QoS policy for the service minus
the policer drops.
BytesOut Aggregate of bytes matching all classes of the egress QoS policy for the service minus the
queuing drops.
PacketsOut Aggregate of packets matching all classes of the egress QoS policy for the service minus
the queuing drops
Dynamic template features that support accounting statistic collection and require that their statistics be
reported in the AAA service accounting records can enable accounting statistics on their features using the
newly-introduced optional acct-stats configuration option. This option is not available for the features that
do not support statistic collection. By default, QoS accounting statistics are disabled to optimize performance.
The QoS counters for each direction is reported only if a QoS policy is applied for that service in the given
direction. For example, if a service does not have an ingress policy applied, BytesIn and PacketsIn counters
are reported as being 0.
Note
Pre-requisites
Subscriber accounting, the parent accounting record for service accounting, must be configured to enable
the service accounting feature to work.
4.3.x
OL-28375-03 59
Service Accounting
The keyword acct-stats must be configured in service-policy configuration to enable the service
accounting feature to report feature counter information as part of the records.
Restriction
IPv4 and IPv6 subscriber sessions has a single set of service accounting records. They are merged into
one set of bytes_in, bytes_out, packets_in, packets_out counters.
Configuring Service Accounting
Perform this task to configure service accounting through the dynamic template:
Before You Begin
You must configure subscriber accounting before performing this task. Refer Creating Dynamic Template
for IPv4 or IPv6 Subscriber Session, on page 84 for configuring procedure.
SUMMARY STEPS
1. configure
2. aaa accounting service {list_name | default} {broadcast group {group_name | radius} |group
{group_name | radius} }
3. aaa service-accounting [extended | brief]
4. dynamic-template
5. type service dynamic-template-name
6. accounting aaa list {method_list_name | default} type service [periodic-interval time]
7. {ipv4 | ipv6} access-group access-list-name
8. service-policy {input | output | type} service-policy_name [acct-stats]
DETAILED STEPS
Example:
Step 1
Creates an accounting list for service
accounting
aaa accounting service {list_name | default} {broadcast group
{group_name | radius} |group {group_name | radius} }
Example:
RP/0/RSP0/CPU0:router(config)# aaa accounting service l1 group
srGroup1
Step 2
Release 4.3.x
60 OL-28375-03
(Optional) Sets accounting parameters for
service to select the level of subscriber
aaa service-accounting [extended | brief]
Example:
RP/0/RSP0/CPU0:router(config)# aaa service-accounting brief
Step 3
accounting state and to identity attribute
reporting in brief or extended form.
The default setting is
extended.
Note
Enters the dynamic-template configuration
mode.
dynamic-template
Example:
RP/0/RSP0/CPU0:router(config)# dynamic-template
Step 4
Creates a dynamic-template with a
user-defined name for a service.
type service dynamic-template-name
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template)# type service s1
Step 5
Configures the service accounting feature. accounting aaa list {method_list_name | default} type service
[periodic-interval time]
Step 6
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template-type)# accounting
aaa list l1 type service periodic-interval 1000
Sets IPv4 or IPv6 access list to an interface. {ipv4 | ipv6} access-group access-list-name
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template-type)# ipv4
access-group ACL1
access-group ACL2
Step 7
Associates a service-policy to the dynamic
template, and enables service accounting
feature using acct-stats keyword.
service-policy {input | output | type} service-policy_name [acct-stats]
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template-type)#
service-policy input QoS1 acct-stats
service-policy output QoS2 acct-stats
Step 8
commitSaves the configuration changes,
and remains within the configuration session.
endPrompts user to take one of these
actions:
YesSaves configuration changes and
exits the configuration session.
4.3.x
OL-28375-03 61
NoExits the configuration session
without committing the configuration
changes.
CancelRemains in the configuration
mode, without committing the
Configuring Service Accounting: Example
configure
aaa accounting service S1 group SG1
aaa service-accounting brief
dynamic-template
type service s1
accounting aaa list S1 type service periodic-interval 600
ipv4 access-group ACL1
service-policy input QOS1 acct-stats
service-policy output QOS2 acct-stats
!
!
end
Statistics Infrastructure
The accounting counters are maintained by the service accounting statistics IDs (statsD) infrastructure. Service
accounting interacts with the statistics infrastructure in this manner:
Each feature has a statistics collector process that is responsible for returning statistics counters for that
feature.
A single collector can handle counters for multiple features.
An accounting process, the service accounting management agent, uses the access library to register for
notifications and request statistics, and pushes to a radius server.
There is a polling period to pull the data from statsD. To support sub-second accuracy on stop records, the
statistics are immediately pulled when the session is terminated, without waiting for any polling method to
get accurate data. The same method is followed by session accounting and service accounting. Sub-second
accuracy is not supported for data reported in interim records, because no data is pulled while sending interim
accounting records.
Configuring Statistics IDs (statsD)
The statsD is configured to poll feature statistics by default every 900 seconds (that is, every 15 minutes).
Perform this task to change the default figure to either increase or decrease the polling interval.
Release 4.3.x
62 OL-28375-03
Statistics Infrastructure
SUMMARY STEPS
1. configure
2. statistics period service-accounting {period | disable}
DETAILED STEPS
Example:
Step 1
Sets collection period for statistics collectors for the service accounting
feature.
statistics period service-accounting {period
| disable}
Example:
RP/0/RSP0/CPU0:router(config)#
Step 2
statistics period service-accounting
1800
Yes Saves configuration changes and exits the configuration
session.
CancelRemains in the configuration mode, without committing
the configuration changes.
Configuring Service Accounting: Example
configure
statistics period service-accounting 1800
end
Understanding Per-VRF AAA Function
The Per VRF AAAfunction allows authentication, authorization, and accounting (AAA) on the basis of virtual
routing and forwarding (VRF) instances. This feature permits the Provider Edge (PE) or Virtual Home Gateway
4.3.x
OL-28375-03 63
Understanding Per-VRF AAA Function
(VHG) to communicate directly with the customer's RADIUS server, (which is associated with the customer's
Virtual Private Network (VPN)), without having to go through a RADIUS proxy.
ISPs must be able to define operational parameters such as AAAserver groups, method lists, systemaccounting,
and protocol-specific parameters, and associate those parameters to a particular VRF instance.
The Per VRF AAAfeature is supported with VRF extensions to server-group, RADIUS, and systemaccounting
commands. The list of servers in server groups is extended to include definitions of private servers, in addition
to references to the hosts in the global configuration. This allows simultaneous access to both customer servers
and global service provider servers. The syntax for the command used to configure per-vrf AAA globally is:
radius source-interface subinterface-name [vrf vrf-name]
RADIUS Double-Dip Feature
BNG supports the RADIUS double-dip feature, where BNG sends the first authentication or authorization
request to a service provider's RADIUS server, which in turn responds with the correct VRF associated with
the subscriber session. Subsequently, the BNG redirects the original request, and sends it as a second request,
to the correct RADIUS server that is associated with the designated VRF.
Additional References
These sections provide references related to implementing RADIUS.
RFCs
Standard/RFC - AAA
Remote Authentication Dial In User Service (RADIUS) RFC-2865
RADIUS Accounting RFC-2866
RADIUS Accounting Modifications for Tunnel Protocol Support RFC-2867
RADIUS Attributes for Tunnel Protocol Support RFC-2868
RADIUS Extensions RFC-2869
IANA Considerations for RADIUS RFC-3575
DSL Forum Vendor-Specific RADIUS Attributes RFC-4679
Dynamic Authorization Extensions to RADIUS RFC-5176
Release 4.3.x
64 OL-28375-03
RADIUS Double-Dip Feature
MIBs
MIBs Link MIBs
To locate and download MIBs for selected platforms,
Cisco IOS releases, and feature sets, use Cisco MIB
Locator found at the following URL:
http://www.cisco.com/go/mibs
Technical Assistance
Link Description
http://www.cisco.com/cisco/web/support/index.html The Cisco Technical Support website contains
thousands of pages of searchable technical content,
including links to products, technologies, solutions,
technical tips, and tools. Registered Cisco.com users
can log in fromthis page to access even more content.
4.3.x
OL-28375-03 65
Release 4.3.x
66 OL-28375-03
C HAP T E R 4
Activating Control Policy
A control policy enables the service provider to define certain actions that are performed during various
subscriber life-cycle events. This chapter provides information about activating control policy on the BNG
router. A control policy is defined using a policy-map. The policy-map contains a set of events - events
during which certain actions are performed. The condition for performing an action is defined in a class-map.
After a class-map is created, it is included in the policy-map. The policy-map is then activated on the router
interface for the policy to take effect. One of the actions that can be performed by the policy map is activating
dynamic template. The dynamic template is a container used to group a set of configuration items to be
applied to a group of subscribers. This chapter covers the following topics:
Control Policy Overview, page 67
Creating Class-Map, page 68
Creating Policy-Map, page 70
Activating Policy-Map, page 74
Defining Dynamic Templates, page 75
Control Policy Overview
Acontrol policy enables the service provider to define actions that must be performed during various subscriber
lifecycle events, such as creation of a session, connectivity loss, and so on. For the complete list of events,
see Control Policy Events, on page 71.
Different actions can be executed for different subscribers based on various match criteria. Some actions that
can be specified in the control policy are:
Authenticating or authorizing a subscriber by an external AAA server
Starting subscriber accounting
Activating specific configurations on the subscriber using dynamic templates
A control policy is deployed using policy-map and class-map. Each policy-map contains a list of events that
the service provider considers applicable to the subscriber lifecycle. The policy-map also defines the actions
that will be performed during these events. However, these actions are performed only when certain conditions
4.3.x
OL-28375-03 67
are met. These conditions are called match criteria. The match criteria are defined in class-maps, which is
included within the policy-map. It is possible to have different match criteria for different subscribers.
For example, a control policy can be created to start the "subscriber authentication" action, when a "session
start" event occurs, for a specific "MAC address" match criteria. After this control policy is deployed, when
the device having the specified MACaddress starts a newsession, BNGinitiates the subscriber authentication
process.
The actions defined in the policy-map are executed by action handlers. For more information about supported
action handlers, see Action Handlers, on page 305.
The following figure shows the structure of control policy. It illustrates that for each policy there can be
multiple events; for each event, there can be multiple classes; and for each class, there can be multiple actions.
As a result, a single policy map can be used to trigger multiple actions, when a match is found for a single or
several criteria, during one or many events.
Figure 4: Control Policy
The following sample configuration shows the control policy structure:
policy-map type control subscriber policy-map-name
event <event-type> [match-all|match-first]
class type control subscriber <class-map-name>
<seq#> <action-type> <action_options>
Creating Class-Map
The class-map is used to define traffic class. The traffic is classified based on match criteria defined in the
class-map. The parameter for match criteria can be protocol, MAC address, input interface, access group, and
so on.
Release 4.3.x
68 OL-28375-03
Creating Class-Map
If more than one match criteria is listed in a single class-map, then instructions must be included defining
how the match criteria are to be evaluated. The evaluation instruction are of two types:
Match-anyA positive match is made if the traffic being evaluated matches any one of the specified
criteria.
Match-allA positive match is made only if the traffic being evaluated matches all specified criteria.
Once a match is made, the traffic is considered as a member of the class.
Each class-map is assigned a name for identification. The class-map name is specified within the policy-map.
For creating a class-map, see Configuring a Class-Map, on page 69.
Configuring a Class-Map
Perform this task to configure a class-map for control policies. As an example, this class-map is created with
the evaluation instruction, "match-any". The match criteria is "protocol" with value "PPP". As a result, a
positive match is made when the session is uses PPP protocol.
SUMMARY STEPS
1. configure
2. class-map type control subscriber match-any class-map-name
3. match protocol ppp
4. end-class-map
DETAILED STEPS
Example:
Step 1
Creates a new subscriber control class-map with a user
defined name.
class-map type control subscriber match-any
class-map-name
Step 2
Example:
RP/0/RSP0/CPU0:router(config)# class-map type
control subscriber match-any clmap1
Enters the class-map mode.
Defines the match evaluation instruction to be "match-any".
Defines the match-criteria to be PPP protocol. match protocol ppp Step 3
Example:
RP/0/RSP0/CPU0:router(config-cmap)# match protocol
ppp
More than one match statement can be applied per
class-map.
Note
4.3.x
OL-28375-03 69
Configuring a Class-Map
Ends the class map configuration. end-class-map
Example:
RP/0/RSP0/CPU0:router(config-cmap)# end-class-map
Step 4
Configuring a Class-Map: An example
class-map type control subscriber match-any DHCP_class
match protocol dhcpv4
end-class-map
!
!
end
Creating Policy-Map
The policy-map is used to define the events for which a set of actions are executed when a match, based on
the class-map definitions, is made. For more information about the supported BNG events, see Control Policy
Events, on page 71.
A policy-map lists a set of events. For each event, a set of class-maps are defined. For each class-map, a series
of actions are listed sequentially. After the policy-map is applied on the BNGrouter interface, when the traffic
matches the criteria mentioned in the class-map, the actions are performed.
If more than one class-map is listed in the policy-map, then instruction has to be specified that defines which
class-maps should be applied. The evaluation instruction are of two types:
First-matchActions are performed only when a match is made for the first class-map.
Match-allActions are performed for all matching classes.
Like with a class-map, each policy-map is assigned a name for identification. The policy-map name is specified
when activating the policy-map on the router interface.
For creating a policy-map, see Configuring a Policy-Map, on page 72.
Release 4.3.x
70 OL-28375-03
Creating Policy-Map
Control Policy Events
Control policy on BNG supports the events listed here. These events need to be defined while creating a
policy-map using the task Configuring a Policy-Map, on page 72.
Session-StartThis event is used by the PPPoE and DHCP access protocols to create a subscriber in
the policy plane. The operator may configure the AAA actions and activate dynamic templates, suitable
for subscriber.
Session-ActivateSome access protocols require a two-stage session bring-up; for example, with PPPoE
subscribers, the PPPoEAccess protocol calls the Session-Start event for first sign of life (FSOL), followed
by Session-Activate during PPP negotiation and authentication. The operator configures the AAAactions
and activates the dynamic templates as suitable for the subscriber.
Service-StopCoA is responsible for generating this event. The BNG operator configures the activate
or deactivate actions, to put the subscriber in a default state when a service is stopped.
Authentication-No-ResponseIf configured, this event is triggered when there is no response from the
AAAserver(s) for an authentication request. This event allows the network access server (NAS) operators
to define how the failure should be handled. If the authentication-no-response event is not configured,
then the authentication failure result is propagated to the access protocol for default handling.
Authorization-No-ResponseIf configured, this event is triggered when there is no response from the
AAA server(s) for an authorization request. This event allows the NAS operators to define how the
failure should be handled. If the authorization-no-response event is not configured, then the authorization
results are propagated to the access protocol for default handling, which causes the client who triggered
the authorization to disconnect the subscriber session.
Authentication-FailureIf configured and if the RADIUS server returns an authentication failure, then
the Policy Rule Engine returns an "Authentication-Success" to the client that originated the request, in
order to prevent it from disconnecting the subscriber. Furthermore, instead of depending on the client
to provide the necessary behavior, the actions within the configured Authentication-Failure event are
applied on the subscriber.
Authorization-FailureThe authorization failure event indicates a RADIUS server rejection for the
access request. If configured, the service provider overrides the default handling of the failure from the
client.
Timed-Policy-ExpiryIf configured, this event is triggered as a result of a policy set-timer action that
is configured and set on a subscriber session. This event allows NAS operators to define a timer for a
number of possible scenarios. The set timer indicates that certain subscriber state changes have taken
place. If sessions are not in the desired state, the NAS operators can disconnect or terminate the session
through a configured disconnect action, or impose a different user policy.
Account-LogonIf configured, this event provides an override behavior to the default account-logon
processing. The default behavior only triggers authentication with provided credentials. However, if
you override the default account-logon event, then you must explicitly configure the authentication
action, and any additional action you require.
Account-LogoffIf configured, this event provides an override behavior for the default account-logoff
processing. The default behavior of the account-logoff processing is to disconnect the subscriber. Being
able to override the default behavior is useful. Instead of disconnecting the subscriber, the service provider
can performa re-authentication. The re-authentication is done through a new account-logon by enabling
HTTP Redirect feature on the subscriber.
4.3.x
OL-28375-03 71
Control Policy Events
Configuring a Policy-Map
Perform this task to configure policy map for control policies. As an example, this policy-map is created for
the Session-Start and Session-Activate events. For the Session-Start event, a dynamic template is activated.
For the Session-Activate event, an authentication process is invoked. For more information about the supported
events, see Control Policy Events, on page 71.
SUMMARY STEPS
1. configure
2. policy-map type control subscriber policy-map-name
3. event session-start match-all
4. class type control subscriber class_name do-until-failure
5. sequence_number activate dynamic-template dynamic-template_name
6. event session-activate match-all
8. sequence_number authenticate aaa list default
9. end-policy-map
DETAILED STEPS
Example:
Step 1
Creates a new policy-map with user-defined name. policy-map type control subscriber policy-map-name Step 2
Example:
RP/0/RSP0/CPU0:router(config)# policy-map type
control subscriber plmap1
Enters the policy-map mode.
Defines an event (session start) for which actions will be
performed.
event session-start match-all
Example:
RP/0/RSP0/CPU0:router(config-pmap)# event
session-start match-all
Step 3
Defines the match instruction to be "match-all", which
executes actions for all matched classes.
Associates a class-map with the event. The class-map name
has to be specified.
class type control subscriber class_name do-until-failure
Example:
RP/0/RSP0/CPU0:router(config-pmap-e)# class type
control subscriber CL1 do-until-failure
Step 4
Instructs that the actions will be performed until a failure
occurs.
Release 4.3.x
72 OL-28375-03
Defines the action to be performed. In this case, it activates
a dynamic-template.
sequence_number activate dynamic-template
dynamic-template_name
Step 5
Example:
RP/0/RSP0/CPU0:router(config-pmap-c)# 1 activate
dynamic-template template1
This command can be repeated to define multiple
actions.
Note
Defines an event (activate session) for which actions will
be performed.
event session-activate match-all
Example:
session-activate match-all
Step 6
Defines the match instruction to be "match-all", which
executes actions for all matched classes.
Associates a class-map with the event. The class-map name
needs to be specified.
Example:
control subscriber CL1 do-until-failure
Step 7
Instructs that the actions will be performed until a failure
occurs.
Defines the action to be performed. In this case, it initiates
the authentication of AAA list.
sequence_number authenticate aaa list default
Example:
RP/0/RSP0/CPU0:router(config-pmap-c)# 2
authenticate aaa list default
Step 8
Ends the policy map configuration. end-policy-map
Example:
RP/0/RSP0/CPU0:router(config-pmap)# end-policy-map
Step 9
Configuring a Policy-Map: An example
policy-map type control subscriber PL1
4.3.x
OL-28375-03 73
event session-start match-first
class type control subscriber DHCP_class do-until-failure
1 activate dynamic-template dhcp
class type control subscriber class-default do-until-failure
! Packet trigger is default
1 activate dynamic-template packet-trigger
end-policy-map
!
!
end
Activating Policy-Map
After a policy-map is created, it needs to be activated on a router interface. The policies are implemented only
after the policy-map activation is completed. One or more policy-maps will constitute the service-policy. To
enable the service-policy, see Enabling a Service-Policy on a Subscriber Interface, on page 74.
Enabling a Service-Policy on a Subscriber Interface
Perform this task to enable a service-policy on a subscriber interface. The process involves attaching a
previously created policy-map with an interface. Once this process is complete, the actions defined in the
class-map will take effect for the traffic coming on the interface on which service-policy is enabled.
SUMMARY STEPS
1. configure
2. interface type interface-path-id
3. service-policy type control subscriber policy_name
DETAILED STEPS
Example:
Step 1
Enters the interface configuration mode for the bundle-ether
access interface.
interface type interface-path-id
Example:
Bundle-Ether100.10
Step 2
For IPoE sessions, it is recommended that Dynamic
ARP learning be disabled in the access-interface,
using the arp learning disable command.
Note
Applies a pre-defined policy-map named 'plmap1' to an access
interface.
service-policy type control subscriber policy_name
Example:
RP/0/RSP0/CPU0:router(config-if)# service-policy
type control subscriber plmap1
Step 3
Release 4.3.x
74 OL-28375-03
Activating Policy-Map
commitSaves the configuration changes, and remains within
NoExits the configuration session without committing
Defining Dynamic Templates
Adynamic template is a container used to group a set of configuration settings, and apply themto the subscriber
sessions. A dynamic template is globally configured through CLI. However, defining the dynamic template
does not immediately cause the configuration to be applied to a subscriber interface. The configuration within
a dynamic template is applied to a subscriber interface, only when the dynamic template is activated using a
control policy. Similarly, the applied configurations are stopped, only when the dynamic template is deactivated
using a control policy.
There are three basic types of dynamic-templates:
PPP templatesIt contains specific configurations related to the PPPoE protocol.
IP Subscriber templatesIt contains specific configurations that are activated on IP subscriber sessions.
Service templatesIt contains service-related configuration that are activated in response to session
life-cycle events. Service templates are precluded fromcontaining interface or media-specific commands.
Adynamic template can either be configured on the CLI, or downloaded fromthe AAAserver. In the following
sample configuration, the policy map activates an IP Subscriber dynamic template that is defined on the CLI.
dynamic-template
type ipsubscriber ipsub
ipv4 unnumbered Loopback400
class type control subscriber class-default do-all
1 activate dynamic-template ipsub
There are two types of dynamic templates that are downloaded from the AAA serveruser profiles and
service profiles. User profiles are applied to a single subscriber, whereas, service profiles can be applied to
multiple subscribers. In the following sample configuration, the policy map downloads a service template
from the AAA server.
Radius Config:
service1 Password="xxxxxx"
Cisco-avpair = "ipv4:ipv4-unnumbered=Loopback400"
4.3.x
OL-28375-03 75
Defining Dynamic Templates
Router Config:
class type control subscriber class-default do-all
1 activate dynamic-template service1 aaa list default
In the above example, the "aaa list default" keyword specifies that the template "service1" be downloaded
from the AAA server. A template is downloaded only once. If there are multiple control policies referring to
service1, then those will get the previously downloaded version.
It is possible to activate more than one dynamic template on the same subscriber interface, for the same event
or different events. If the configurations for a particular functionality is defined in multiple dynamic templates,
the configurations are derived from all the templates on a certain order of precedence. This order is based on
the type of dynamic template, and whether it is being applied from CLI or AAA. The order is:
Template applied by the user profile from AAA
Template applied by the service profile from AAA
IP Subscriber template applied from CLI
PPP template applied from CLI
Service template applied from CLI
The tasks involving the use of dynamic templates to define specific feature configurations are included in
their corresponding feature topics.
These sections provide references related to implementing control policy.
MIBs
MIBs Link MIB
Release 4.3.x
76 OL-28375-03
Link Description
http://www.cisco.com/cisco/web/support/index.html The Cisco Support website provides extensive online
resources, including documentation and tools for
troubleshooting and resolving technical issues with
Cisco products and technologies.
To receive security and technical information about
your products, you can subscribe to various services,
such as the Product Alert Tool (accessed from Field
Notices), the Cisco Technical Services Newsletter,
and Really Simple Syndication (RSS) Feeds.
Access to most tools on the Cisco Support website
requires a Cisco.com user ID and password.
4.3.x
OL-28375-03 77
Release 4.3.x
78 OL-28375-03
C HAP T E R 5
A subscriber accesses network resources through a logical connection known as subscriber session. This
chapter provides information about various types of subscriber sessions, namely IPoE and PPPoE, and IP
addressing by DHCP. This chapter covers these topics:
Subscriber Session Overview, page 79
Establishing IPoE Session, page 81
Establishing PPPoE Session, page 91
Making DHCP Settings, page 122
DHCPv6 Overview, page 136
Packet Handling on Subscriber Interfaces, page 168
IPv6 Neighbor Discovery, page 170
Subscriber Session Overview
A session represents the logical connection between the customer premise equipment (CPE) and the network
resource. To enable a subscriber access the network resources, the network has to establish a session with the
subscriber. Each session establishment comprises of these phases:
4.3.x
OL-28375-03 79
When packets arrive on an access interface, an attempt is made to link that packet to a subscriber context. Note
For PPPoE sessions the Source MAC of the CPE, Access interface and PPPoE Session ID are used
to match the remote peer to a subscriber interface.
For IPoE sessions the Source MAC, Access interface and IP address are verified against the DHCP
binding to find a matching subscriber interface.
If there is no match, the packet is mapped against the access (sub-)interface. Considering that the access
interface in IPoE designs is IP enabled (eg via an IP-Unnumbered configuration) that packets are processed
like regular IP. In order to secure your BNG access interface, you will want to apply either uRPF or an
Access-List blocking everything but DHCP incoming on the access interface to limit remote subscribers
for which we don't have an interface created from accessing network resources.
Establishing a connectionin this phase CPE finds the BNG with which to communicate.
Authenticating and authorizing the subscriberin this phase, BNG authenticates the subscribers and
authorizes them to use the network. This phase is performed with the help of the RADIUS server.
Giving subscriber an identityin this phase, the subscriber is assigned an identity, the IP address.
Monitoring the sessionin this phase, BNG ascertains that the session is up and running.
The subscribers are not configured directly on BNG. Instead, a framework is created on which subscriber
features and subscriber sessions are started and stopped dynamically. The framework consists of control
policies and dynamic templates, which perform these functions:
Control policy determines the action BNG takes when specific events, such as receipt of a session start
request, or failure of authentication, occurs. The action is determined by the class-map defined in the
control policy. The action involves activating dynamic templates.
Dynamic template contains a set of CLI commands that are applied to a subscriber session. Multiple
dynamic templates can be activated, one at a time, on the same subscriber interface. Also, the same
dynamic template can be activated on multiple subscriber interfaces through different control policies.
Service providers can deploy subscribers over VLAN in these ways:
1:1 VLAN modelThis model depicts a scenario where one dedicated VLAN is available for each
customer. Each VLAN is an q-in-q VLAN where the inner VLAN tag represents the subscriber and the
outer VLAN tag represents the DSLAM.
N:1 VLAN modelThis model depicts a scenario where multiple subscribers are available on a shared
VLAN. The VLAN tags represent the DSLAM or the aggregation device.
Ambiguous VLANs This model allows the operator to specify a large number of VLANs in a single
CLI line. Using ambiguous VLAN, a range of inner or outer tags (or both) can be configured on a VLAN
sub-interface. This is particularly useful for the 1:1 model, where every subscriber has a unique value
for the set of VLAN tags. For more information about ambagious VLANs, see Subscriber Session on
Ambiguous VLANs , on page 230.
The subscriber sessions are established over the subscriber interfaces, which are virtual interfaces. It is possible
to create only one interface for each subscriber session. A port can contain multiple VLANs, each of which
can support multiple subscribers. BNG creates subscriber interfaces for each kind of session. These interfaces
Release 4.3.x
80 OL-28375-03
Subscriber Session Overview
are named based on the parent interface, such as bundle-ether 2.100.pppoe312. The subscribers on bundles
(or bundle-VLANs) interfaces allow redundancy, and are managed on the BNG route processor (RP).
To provide network redundancy and load balancing, the service provider can deploy multiple links between
the DSLAM and the BNG. The individual links can be grouped into ether-bundles, including VLANs over
ether-bundles, or link aggregation groups (LAGs). The subscriber sessions can be active on any link within
the bundle or group. If a BNG is deployed in a LAG configuration, all traffic for one subscriber should be
configured to traverse one link of the ether-bundle. Load-balancing is achieved by putting different subscribers
on different links.
There are two mechanisms to establish a subscriber session, namely, IPoE and PPPoE. These are discussed
next in the next topics.
If a clear subscriber session all command is issued with the intend to clear all the subscriber sessions
and if a route processor fail over (RPFO) occurs while the session bring down is in progress, then it is
recommended to re-run the same command post RPFO, to ensure all the remaining sessions, if any, are
brought down.
Note
Restrictions
On BNG, only dynamic creation of subscribers is supported. Also, the subscriber must be present only
on the bundle interfaces.
If the subscriber's VRF is taken from the access interface's VRF value, then the VRF, configured in the
dynamic template used by the subscriber, must match. If the two VRFs do not match, then the session
would not work properly.
Establishing IPoE Session
In an IPoE subscriber session, subscribers run IPv4 or IPv6 on the CPE device and connect to the BNGthrough
a Layer-2 aggregation or Layer-3 routed network. IP subscriber sessions that connect through a Layer-2
aggregation network are called L2-connected. IPoE subscriber sessions are always terminated on BNG and
then routed into the service provider network. IPoE relies on DHCP to assign IP address. A typical IPoE
session is depicted in the following figure.
Figure 5: IPoE Session
4.3.x
OL-28375-03 81
Establishing IPoE Session
The process of provisioning an IPoE session involves:
Enabling the processing of IPv4 or IPv6 protocol on an access interface. See Enabling IPv4 or IPv6 on
an Access Interface, on page 82.
For subscriber deployments, it is recommended that Dynamic ARP learning be disabled
in the access-interface, using the arp learning disable command in the access-interface
configuration mode.
Note
Creating dynamic template that contains the settings for the IPoE sessions. See Creating Dynamic
Template for IPv4 or IPv6 Subscriber Session, on page 84.
Creating policy-map to activate dynamic template. See Creating a Policy-Map to Run During IPoE
Session, on page 87.
Enabling IPoE subscriber creation on access interface by activating service-policy. The service-policy
will apply the policy-map on the access interface. See Enabling IPoE Subscribers on an Access Interface,
on page 88.
If an access interface in BNGis configured to support only packet (PKT) triggered sessions, or both DHCP
and PKT triggered sessions, then a burst of traffic with unique flows can affect the BNG router in terms
of processing each packet to determine if it is an IPoE (PKT triggered) packet. New subscriber sessions
cannot be established in these scenarios and this can in turn lead to system instability. Therefore, it is
mandatory to configure static lpts policer for unclassified rsp protocol, on each of the line cards (LCs),
such that the traffic rate does not exceed 150 pps per LC. The rate configured is applied at network
processor (NP). Therefore, for an LC with 4 NPs, the rate should be configured as 38 (150/4), to achieve
a traffic rate of 150 pps. For example, lpts punt police location 0/RSP0/CPU0 protocol unclassified
rsp rate 38.
Note
Enabling IPv4 or IPv6 on an Access Interface
Perform these tasks to enable IPv4 and IPv6 processing on an access interface. In this example, the IPv4 is
being provisioned on an unnumbered bundle-interface.
SUMMARY STEPS
1. configure
3. arp learning disable
4. ipv4 unnumbered interface-type interface-instance
5. ipv6 enable
Release 4.3.x
82 OL-28375-03
DETAILED STEPS
Example:
Step 1
Enters interface configuration mode for the bundle-interface. interface type interface-path-id
Example:
interface Bundle-Ether100.10
Step 2
Disables arp learning for the access-interface. arp learning disable
Example:
RP/0/RSP0/CPU0:router(config-if)# arp
learning disable
Step 3
Enables IPv4 processing on a unnumbered interface without assigning
an explicit IPv4 address to that interface. Instead, the IP address is
ipv4 unnumbered interface-type
interface-instance
Step 4
borrowed from the loopback interface. For the "ipv4 unnumbered"
Example:
RP/0/RSP0/CPU0:router(config-if)# ipv4
unnumbered loopback 5
command, you must specify another interface in the same router that has
been assigned an IP address and whose status is displayed as up for the
show interfaces command.
Enables IPv6 processing on an unnumbered interface that has not been
assigned an explicit IPv6 address.
ipv6 enable
Example:
RP/0/RSP0/CPU0:router(config-if)# ipv6
enable
Step 5
This step not only enables IPv6 processing on the interface, but
also assigns an IPv6 link-local unicast address to it.
Note
session.
4.3.x
OL-28375-03 83
Enabling IPv4 or IPv6on an Access Interface: Examples
//Enabling IPv4 on an Access Interface
configure
arp learning disable
ipv4 unnumbered loopback 5
!
!
end
//Enabling IPv6 on an Access Interface
configure
ipv6 enable
!
!
end
Creating Dynamic Template for IPv4 or IPv6 Subscriber Session
Perform this task to create a dynamic template for IPv4 or IPv6 subscriber session. As an example, in this
dynamic template you will specify the MTU value for the IPv4 or IPv6 session and enable uRPF. The uRPF
ensures that the traffic from malformed or forged IPv4 source addresses are not accepted on the subscriber
interface. For more information about uRPF feature, see uRPF, on page 232.
SUMMARY STEPS
1. configure
2. dynamic-template
3. type { ipsubscriber | ppp | service } dynamic-template-name
4. timeout idle value
5. accounting aaa list default type session periodic-interval value dual-stack-delay value
6. {ipv4 | ipv6} mtu mtu-bytes
7. {ipv4 | ipv6}verify unicast source reachable-via rx
DETAILED STEPS
Example:
Step 1
Release 4.3.x
84 OL-28375-03
Enters the dynamic-template configuration mode. dynamic-template
Example:
Step 2
Creates a dynamic-template with an user-defined
name for IP subscriber.
type { ipsubscriber | ppp | service } dynamic-template-name
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template)# type
ipsubscriber ipsub1
Step 3
IPv4 or IPv6 or Dual-stack Subscribers support
idle timeout feature.
timeout idle value
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template-type)# timeout
idle 600
Step 4
Note The idle timeout is supported only
for ipsubscriber and service type
templates.
In Release 4.2.3, if the idle timeout
is configured, it is expected that the
IPoE sessions terminate when the
timer configured under
dynamic-template or the timer
passing down from the RADIUS
server expires. Whereas in Release
4.2.1, the IPoE and PPPoE sessions
are retained idle when the idle
timeout period expires.
Configures the subscriber accounting feature. accounting aaa list default type session periodic-interval value
dual-stack-delay value
Step 5
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template)# accounting
aaa list default type session periodic-interval 60
dual-stack-delay 1
Sets IPv4 or IPv6 maximum transmission unit
(MTU). The range is from68 to 65535 bytes. The
{ipv4 | ipv6} mtu mtu-bytes
Example:
mtu 678
mtu 548
Step 6
MTU value defines the largest packet size that
can be transmitted during the subscriber session.
4.3.x
OL-28375-03 85
Enables uRPF for packet validation that performs
source address reachability check.
{ipv4 | ipv6}verify unicast source reachable-via rx
Example:
verify unicast source reachable-via rx
verify unicast source reachable-via rx
Step 7
YesSaves configuration changes and exits
mode, without committing the configuration
changes.
Creating Dynamic Template for IPv4 or IPv6 Subscriber Session: Examples
//Creating Dynamic Template for IPv4 Subscriber Session
configure
dynamic-template
type ipsubscriber ipsub1
timeout idle 600
accounting aaa list default type session periodic-interval 60 dual-stack-delay 1
ipv4 mtu 678
ipv4 verify unicast source reachable-via rx
!
!
end
//Creating Dynamic Template for IPv6 Subscriber Session
configure
dynamic-template
type ipsubscriber ipsub1
timeout idle 600
accounting aaa list default type session periodic-interval 60 dual-stack-delay 1
ipv6 mtu 678
ipv6 verify unicast source reachable-via rx
!
!
end
Release 4.3.x
86 OL-28375-03
Creating a Policy-Map to Run During IPoE Session
Perform this task to create a policy-map that will activate a predefined dynamic-template during an IPoE
subscriber session. As an example, this policy-map activates a dynamic template, and applies a locally defined
authorization setting, during a session-start event.
SUMMARY STEPS
1. configure
2. policy-map type control subscriber policy_name
3. event session-start match-first
6. sequence_number authorize aaa list default format format_name password password
DETAILED STEPS
Example:
Step 1
Creates a new policy map of the type "control subscriber" with
the name "IPoE_policy".
policy-map type control subscriber policy_name
Example:
control subscriber IPoE_policy
Step 2
performed.
Example:
session-start match-first
Step 3
Configures the class to which the subscriber has to be matched.
When there is a match, executes all actions until a failure is
encountered.
class type control subscriber class_name
do-until-failure
Example:
RP/0/RSP0/CPU0:router(config-pmap-e)# class
Step 4
type control subscriber class-default
do-until-failure
4.3.x
OL-28375-03 87
Creating a Policy-Map to Run During IPoE Session
Allows authentication of the subscriber to be triggered using
the complete structure username.
Example:
activate dynamic-template ipsub1
Step 5
Allows authorization of the subscriber to be triggered using the
domain name of the subscriber. Also, provides domain
sequence_number authorize aaa list default format
format_name password password
Step 6
format-rule, which helps to parse the domain from a complete
structured username.
Example:
authorize aaa list default format RM_User
password Cisco
Creating a Policy-Map to Run During IPoE Session: An example
configure
policy-map type control subscriber IPoE_policy
1 activate dynamic-template ipsub1
1 authorize aaa list default format RM_User password Cisco
!
!
end
Enabling IPoE Subscribers on an Access Interface
Perform this task to enable IPoE subscriber creation on an access interface.
Release 4.3.x
88 OL-28375-03
SUMMARY STEPS
1. configure
2. interface interface-type interface-path-id
3. arp learning disable
4. {ipv4 |ipv6} address {ipv4_address |ipv6_address} ipsubnet_mask
5. service-policy type control subscriber policy-name
6. encapsulation dot1q value
7. ipsubscriber {ipv4 |ipv6}l2-connected
8. initiator dhcp
9. initiator unclassified-source
DETAILED STEPS
Example:
Step 1
Configures an interface and enters interface
configuration mode.
interface interface-type interface-path-id
Example:
Bundler-Ether400.12
Step 2
The type argument specifies an interface type. For
more information on interface types, use the
question mark (?) online help function.
The instance argument specifies either a physical
interface instance or a virtual instance.
The naming notation for a physical interface
instance is rack/slot/module/port. The slash
(/) between values is required as part of the
notation.
The number range for a virtual interface
instance varies depending on the interface
type.
Disables arp learning for the access-interface. arp learning disable
Example:
RP/0/RSP0/CPU0:router(config-if)# arp learning disable
Step 3
Sets the IPv4 address or an IPv6 address for an interface. {ipv4 |ipv6} address {ipv4_address |ipv6_address}
ipsubnet_mask
Step 4
4.3.x
OL-28375-03 89
Example:
RP/0/RSP0/CPU0:router(config-subif)# ipv4 address
3.5.1.1 255.255.0.0
or
Example:
RP/0/RSP0/CPU0:router(config-subif)# ipv6 address
1144:11
Associates a subscriber control service policy to the
interface.
service-policy type control subscriber policy-name
Example:
RP/0/RSP0/CPU0:router(config-subif)# service-policy
type control subscriber PL4
Step 5
Refer to the "Configuring a Subscriber Policy
Map" procedure to create a PL4 policy-map.
Note
Defines the matching criteria to map 802.1Q frames
ingress on an interface to the appropriate service
instance. The value ranges from 1 to 4094.
encapsulation dot1q value
Example:
RP/0/RSP0/CPU0:router(config-subif)# encapsulation
dot1q 40
Step 6
Enables creations of L2-connected IPv4 or IPv6
subscribers on the sub-interface.
ipsubscriber {ipv4 |ipv6}l2-connected
Example:
RP/0/RSP0/CPU0:router(config-subif)# ipsubscriber ipv4
l2-connected
Step 7
or
Example:
RP/0/RSP0/CPU0:router(config-subif)# ipsubscriber ipv6
l2-connected
Enables DHCP sourced IPoE subscriber creation on the
access-interface that can also be combined with
unclassified-source initiator.
initiator dhcp
Example:
RP/0/RSP0/CPU0:router(config-subif-ipsub-ipv4-l2conn)#
initiator dhcp
Step 8
or
Example:
initiator dhcp
Initiates a class for DHCP for an IP session initiated by
DHCP.
initiator unclassified-source
Example:
Step 9
If multiple initiators are used, use a policy or
class map to prevent overlap of the IP addresses
for the different sources.
Note
or
Example:
Release 4.3.x
90 OL-28375-03
Enabling IPoE Subscribers on an Access Interface: Examples
configure
interface Bundler-Ether400.12
ipv4 address 3.5.1.1 255.255.0.0
service-policy type control subscriber PL4
encapsulation dot1q 40
ipsubscriber ipv4 l2-connected
initiator dhcp
!
!
end
configure
interface Bundler-Ether400.12
ipv6 address 4444:34
initiator dhcp
!
!
end
Establishing PPPoE Session
The PPP protocol is mainly used for communications between two nodes, like a client and a server. The PPP
protocol provides a standard method for transporting multi-protocol diagrams over point-to-point links. It
defines an encapsulation scheme, a link layer control protocol (LCP), and a set of network control protocols
(NCPs) for different network protocols that can be transmitted over the PPP link. The LCP is used to configure
and maintain the data link. PPP peers can use the LCP to negotiate various link layer properties or
characteristics. The NCP is used to establish and configure the associated network protocol before data packets
for the protocol can be transmitted.
4.3.x
OL-28375-03 91
Establishing PPPoE Session
One of the methods to establish PPP connection is by the use of PPP over Ethernet (PPPoE). In a PPPoE
session, the Point-to-Point (PPP) protocol runs between the CPE and BNG. The Home Gateway (which is
part of the CPE) adds a PPP header (encapsulation) that is terminated at the BNG.
CPE detects and interacts with BNG using various PPPoE Active Discovery (PAD) messages listed here:
PPPoE Active Discovery Initiation (PADI)The CPE broadcasts to initiate the process to discover
BNG.
PPPoE Active Discovery Offer (PADO)The BNG responds with an offer.
PPPoE Active Discovery Request (PADR)The CPE requests to establish a connection.
PPPoE Active Discovery Session confirmation (PADS)BNG accepts the request and responds by
assigning a session identifier (Session-ID).
PPPoE Active Discovery Termination (PADT)Either CPE or BNG terminates the session.
In redundant BNG setups, where the PPPoE client is connected to multiple BNGs, the PADI message sent
by the CPE is received on all BNGs. Each BNG, in turn, replies with a PADO message. You must configure
Smart Server Selection on BNG to allow subscribers to select one of the BNGs in a multi-BNG setup. Refer
PPPoE Smart Server Selection, on page 116
The BNGprovides configuration flexibility to limit and throttle the number of PPPoE sessions requests, based
on various parameters. For details, see PPPoE Session Limit, on page 119 and PPPoE Session Throttle, on
page 121.
The PPPoE session are of two types, PPP PTA and PPP LAC. For the functioning of PPP PTA and PPP LAC
session, the RADIUS server must be set up to authenticate and forward sessions as necessary. There is no
local authentication available on BNG. The PPP PTA and PPP LAC sessions are explained in the sections,
Provisioning PPP PTA Session, on page 92 and Provisioning PPP LAC Session, on page 99.
Provisioning PPP PTA Session
In a PPP Termination and Aggregation (PTA) session, the PPP encapsulation is terminated on BNG. After it
is terminated, BNGroutes the traffic to the service provider using IP routing. Atypical PTAsession is depicted
in this figure.
Figure 6: PTA Session
PPPoE session configuration information is contained in PPPoE profiles. After a profile has been defined, it
can be assigned to an access interface. Multiple PPPoE profiles can be created and assigned to multiple
interfaces. A global PPPoE profile can also be created; the global profile serves as the default profile for any
interface that has not been assigned a specific PPPoE profile.
Release 4.3.x
92 OL-28375-03
The PPP PTA session is typically used in the Network Service Provider (retail) model where the same service
operator provides the broadband connection to the subscriber and also manages the network services. The
process of provisioning a PPP PTA session involves:
Creating a PPPoE profile for PPPoE session. See, Creating PPPoE Profiles, on page 93.
Creating dynamic template that contains the various settings for the PPPoE sessions. See, Creating a
PPP Dynamic-Template, on page 94.
Creating policy-map to activate the dynamic template. See, Creating a Policy-Map to Run During PPPoE
Session, on page 95.
Enabling subscriber creation, and apply the PPPoE profile and service-policy on the access interface.
See, Applying the PPPoE Configurations to an Access Interface, on page 97.
The subscriber creation function must be explicitly enabled on BNG. Unless this function is enabled,
the system will not attempt subscriber classification. As a result, the packets get forwarded based on the
incoming interface mode.
Creating PPPoE Profiles
Perform this task to create PPPoE profiles. The PPPoE profile will later be applied to an access interface.
SUMMARY STEPS
1. configure
2. pppoe bba-group bba-group name
3. service name service_name
DETAILED STEPS
Example:
Step 1
Creates a PPPoE profile with an user-specified name. pppoe bba-group bba-group name
Example:
RP/0/RSP0/CPU0:router(config)# pppoe bba-group
bba_1
Step 2
Indicates the service that is requested by the subscriber. service name service_name Step 3
Example:
RP/0/RSP0/CPU0:router(config-bbagroup)# service
name service_1
Repeat this step for each service name that you want to add to
the subscriber profile.
4.3.x
OL-28375-03 93
Creating PPPoE Profiles: An example
configure
pppoe bba-group bba_1
service name service_1
!
!
end
Creating a PPP Dynamic-Template
Perform this task to create a PPP dynamic-template. As an example, this dynamic-template is created to apply
PAP and CHAP authentication methods.
SUMMARY STEPS
1. configure
2. dynamic-template type ppp dynamic_template_name
3. ppp authentication pap
4. ppp authentication chap
DETAILED STEPS
Example:
Step 1
Release 4.3.x
94 OL-28375-03
Creates a dynamic-template with user-defined name
for PPP session.
dynamic-template type ppp dynamic_template_name
Example:
RP/0/RSP0/CPU0:router(config)# dynamic-template type ppp
ppp_pta_template
Step 2
Enables the use of PAP type authentication during
link negotiation by Link Control Protocol (LCP).
ppp authentication pap
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template-type)# ppp
authentication pap
Step 3
Enables the use of CHAP type authentication during
link negotiation by Link Control Protocol (LCP).
ppp authentication chap
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template-type)# ppp
authentication chap
Step 4
Creating a PPP Dynamic-Template: An example
configure
dynamic-template type ppp ppp_pta_template
ppp authentication pap
ppp authentication pap chap
!
!
end
Creating a Policy-Map to Run During PPPoE Session
Performthis task to create a policy-map that will activate a PPP dynamic-template during a PPPoE subscribers
session. As an example, this policy-map activates a dynamic template during a session-start event. Also, this
policy-map applies a locally-defined authorization setting during a session-activate event.
4.3.x
OL-28375-03 95
SUMMARY STEPS
1. configure
6. event session-activate match-all
DETAILED STEPS
Example:
Step 1
Creates a new policy map of the type "control subscriber"
with the user-defined name "PPPoE_policy".
policy-map type control subscriber policy_name
Example:
control subscriber PPPoE_policy
Step 2
performed.
Example:
Step 3
Configures the class to which the subscriber is to be matched.
encountered.
do-until-failure
Example:
control subscriber pta_class do-until-failure
Step 4
Activates the dynamic-template with the specified dynamic
template name.
Example:
dynamic-template ppp_pta_template
Step 5
Release 4.3.x
96 OL-28375-03
Defines an event (session activate) for which actions are
performed.
Example:
session-activate match-all
Step 6
Configures the class to which the subscriber is to be matched.
encountered.
do-until-failure
Example:
control subscriber PPP_class do-until-failure
Step 7
Allows authentication of the subscriber to be triggered using
the complete structure username.
Example:
Step 8
Creating a Policy-Map to Run During PPPoE Session: An example
configure
policy-map type control subscriber policy1
class type control subscriber pta_class do-until-failure
1 activate dynamic-template template1
!
!
class type control subscriber pta_class1 do-until-failure
1 activate dynamic-template ppp_pta_template
end-policy-map
Applying the PPPoE Configurations to an Access Interface
Perform this task to apply the PPPoE profiles and the policy-maps to an access interface. The completion of
this task enables the receiving of PPPoE traffic on the interface.
4.3.x
OL-28375-03 97
Before You Begin
You must perform this task after performing the Creating PPPoE Profiles, on page 93.
SUMMARY STEPS
1. configure
4. pppoe enable bba-group bbagroup_name
DETAILED STEPS
Example:
Step 1
Enters interface configuration mode for the bundle-interface. interface type interface-path-id
Example:
Bundle-Ether 5.1
Step 2
Associates a subscriber control service policy to the interface. service-policy type control subscriber policy_name
Example:
RP/0/RSP0/CPU0:router(config-if)# service-policy
type control subscriber PL1
Step 3
Enables PPPoE on the bundle-ether interface and specifies
the PPPoE profile named bba_1 to be used on this interface.
pppoe enable bba-group bbagroup_name
Example:
RP/0/RSP0/CPU0:router(config-if)# pppoe enable
bba-group bba_1
Step 4
Release 4.3.x
98 OL-28375-03
Applying the PPPoE Configurations to an Access Interface: An example
configure
pppoe enable bba-group bba_1
!
!
end
Provisioning PPP LAC Session
In a PPP LACsession, the PPP session is tunneled to a remote network server by BNG, using Layer 2 Tunneling
Protocol (L2TP). BNGperforms the role of L2TP Access Concentrator (LAC), as it puts the subscriber session
in the L2TP tunnel. The device on which the tunnel terminates is called L2TP Network Server (LNS). During
a PPP LAC session, the PPPoE encapsulation terminates on BNG; however, the PPP packets travel beyond
BNG to LNS through the L2TP tunnel. A typical LAC session is depicted in Figure 1.
Figure 7: LAC Session
The PPP LAC session is used in the Access Network Provider (wholesale) model, where the network service
provider (NSP) is a separate entity from the local access network provider (ANP). NSPs perform access
authentication, manage and provide IP addresses to subscribers, and are responsible for overall service. The
ANP is responsible for providing the last-mile digital connectivity to the customer, and for passing on the
subscriber traffic to the NSP. In this kind of setup, the ANP owns the LAC and the NSP owns the LNS.
A PPP LAC session establishes a virtual point-to-point connection between subscriber device and a node in
the service provider network. The subscriber dials into a nearby L2TP access connector (LAC). Traffic is then
securely forwarded through the tunnel to the LNS, which is present in service provider network. This overall
deployment architecture is also known as Virtual Private Dial up Network (VPDN).
Reassembly of fragmented L2TP data packets is enabled on LAC to prevent these packets from getting
dropped. See, L2TP Reassembly on LAC, on page 100
A PPP LAC session supports stateful switchover (SSO) along with non-stop routing (NSR) to reduce traffic
loss during RP failover. For more information, see L2TP Access Concentrator Stateful Switchover, on page
102
4.3.x
OL-28375-03 99
The process of provisioning a PPP LAC session involves:
Defining a template with specific settings for the VPDN. See, Configuring the VPDN Template, on
page 105.
Defining the maximum number of VPDN sessions that can be established simultaneously. See,
Configuring Maximum Simultaneous VPDN Sessions, on page 107.
Activating the logging of VPDN event messages. See, Activating VPDN Logging, on page 109.
Specifying the method to apply calling station-ID. See, Configuring Options to Apply on Calling Station
ID, on page 110.
Specifying the session-ID. See, Configuring L2TP Session-ID Commands, on page 111.
Defining specific settings for the L2TP class. See, Configuring L2TP Class Options, on page 113.
Preventing creation of additional VPDN sessions. See, Configuring Softshut for VPDN, on page 115.
Restriction
Provisioning PPP LACsession is subjected to a restriction that only ASR9000 Enhanced Ethernet Line Cards
are supported as core facing line cards.
L2TP Reassembly on LAC
The L2TP Reassembly feature on L2TP Access Concentrator (LAC) ensures reassembly of fragmented L2TP
data packets in the intervening network, between the LAC and L2TP Network Server (LNS). Data packets
are fragmented when they exceed the Maximum Transmission Unit (MTU) of the IPv4 core. Enabling this
feature prevents the fragmented packets from getting dropped and ensures the subsequent forwarding of these
data packets.
When L2TP Reassembly feature is disabled on LAC, fragmented data packets are dropped. The feature does
not affect the reassembly of non-L2TP packets. To ensure that packets for non-L2TP applications are properly
reassembled regardless of whether load balancing occurs for each packet, it is recommended that:
A separate loopback address be configured only for L2TP traffic. No other applications on the router
should use this IP address.
Multiple loopback addresses be used for L2TP, but no other applications across all VRFs should use
these addresses.
In cases of reassembly errors or fragmentation timeout, the maximum period a traffic flow is kept, before it
is forwarded to the Route Switch Processor (RSP) is 250ms.
Restrictions
Enabling L2TP reassembly feature is subjected to these restrictions:
Only ASR 9000 Enhanced Ethernet Line Cards support L2TP reassembly feature
Only IPv4 fragmented packets are reassembled
Only packets with two fragments are reassembled
The fragments must not overlap
The fragmented IP headers must not contain options
Release 4.3.x
100 OL-28375-03
The fragmented L2TP packets must be delivered to the same line card. In other words, the intervening
network must not use per packet load balancing schemes that make the fragments arrive on different
line cards. On the other hand, the reassembly of non-L2TP packets is not affected even when the packets
arrive on different line cards
Enabling L2TP Reassembly on LAC
Perform this task to enable L2TP reassembly on L2TP Access Concentrator (LAC):
SUMMARY STEPS
1. configure
2. vpdn
3. l2tp reassembly
DETAILED STEPS
Example:
Step 1
Enters the VPDN configuration mode. vpdn
Example:
RP/0/RSP0/CPU0:router(config)# vpdn
Step 2
Enables L2TP reassembly on LAC. l2tp reassembly
Example:
RP/0/RSP0/CPU0:router(config-vpdn)# l2tp
reassembly
Step 3
session.
4.3.x
OL-28375-03 101
Enabling L2TP Reassembly on LAC: An example
configure
vpdn
l2tp reassembly
!
end
L2TP Access Concentrator Stateful Switchover
The L2TP Access Concentrator Stateful Switchover (LAC SSO) feature establishes one of the RPs as the
active processor, designates the other RP as the standby processor, and then synchronizes critical state
information between them. In specific Cisco networking devices that support dual RPs, LAC SSO takes
advantage of RP redundancy to increase network availability.
LAC SSO supports non-stop routing (NSR) for VPDN and L2TP protocols in the event of a RP failover. The
NSR provides the ability to guarantee reliable L2TP and VPDN synchronization between active and standby
RPs. In case of RP fail-over, all VPDN and L2TP tunnels and sessions information are preserved without
impacting the L2TP network peer. Also, peer networking devices do not experience routing flaps, and therefore
reduce loss of service outages for customers. When VPDN and LAC SSO are enabled, all the tunnels and
sessions are mirrored to the backup RP.
Enabling LAC SSO
Perform this task to enable LAC/VPDN SSO feature:
SUMMARY STEPS
1. configure
2. vpdn
3. redundancy
5. show vpdn redundancy
6. show vpdn redundancy mirroring
7. show l2tpv2 redundancy
8. show l2tpv2 redundancy mirroring
DETAILED STEPS
Example:
Step 1
Release 4.3.x
102 OL-28375-03
Enters vpdn configuration mode. vpdn
Example:
Step 2
Enters vpdn redundancy configuration mode. redundancy
Example:
RP/0/RSP0/CPU0:router(config-vpdn)#
redundancy
Step 3
Displays all vpdn redundancy related information. show vpdn redundancy
Example:
RP/0/RSP0/CPU0:router# show vpdn redundancy
Step 5
Displays vpdn related mirroring statistics. show vpdn redundancy mirroring
Example:
RP/0/RSP0/CPU0:router# show vpdn redundancy
mirroring
Step 6
Displays L2TP redundancy related information. show l2tpv2 redundancy
Example:
RP/0/RSP0/CPU0:router# show l2tpv2 redundancy
Step 7
Displays L2TP related mirroring statistics. show l2tpv2 redundancy mirroring
Example:
RP/0/RSP0/CPU0:router# show l2tpv2 redundancy
mirroring
Step 8
4.3.x
OL-28375-03 103
Enabling LAC SSO: Example
configure
vpdn
redundancy
process-failures switchover
end
Enabling RPFO on Process-failures
In the event of an application or process crash, if VPDN NSR is enabled, an RP failover is triggered and a
new primary RP process restarts without traffic loss.
The VPDN NSR is disabled by default. Perform these steps to enable RPFO:
SUMMARY STEPS
1. configure
2. nsr process-failures switchover
3. vpdn
4. redundancy
5. process-failures switchover
DETAILED STEPS
Example:
Step 1
Enables VPDN non-stop routing. nsr process-failures switchover
Example:
RP/0/RSP0/CPU0:router(config)# l2tp nsr
Step 2
Enters vpdn configuration mode. vpdn
Example:
Step 3
Enters vpdn redundancy configuration mode. redundancy
Example:
RP/0/RSP0/CPU0:router(config-vpdn)# redundancy
Step 4
Release 4.3.x
104 OL-28375-03
Forces a switchover in case of a process failure. process-failures switchover
Example:
RP/0/RSP0/CPU0:router(config-vpdn-redundancy)#
Step 5
Configuring the VPDN Template
Perform this task to configure the vpdn template:
SUMMARY STEPS
1. configure
2. vpdn template
3. l2tp-class class_name
4. tunnel busy timeout timeout_value
5. caller-id mask-method remove match match_substring
6. dsl-line-info-forwarding
7. ip tos type_of_service_value
8. vpn id value
9. vpn vrf vrf_name
4.3.x
OL-28375-03 105
DETAILED STEPS
Example:
Step 1
Enters the VPDN template sub-mode. vpdn template
Example:
RP/0/RSP0/CPU0:router(config)# vpdn template
Step 2
Configures the l2tp class command. l2tp-class class_name
Example:
RP/0/RSP0/CPU0:router(config-vpdn-template)# l2tp-class class_temp
Step 3
Configure l2tp tunnel busy list
commands. The busy timeout value
ranges from 60-65535.
tunnel busy timeout timeout_value
Example:
RP/0/RSP0/CPU0:router(config-vpdn-template)# tunnel busy timeout 456
Step 4
Configures options to apply on calling
station id by masking the characters
by the match substring specified.
caller-id mask-method remove match match_substring
Example:
RP/0/RSP0/CPU0:router(config-vpdn-template)# caller-id mask-method
remove match m1
Step 5
Forwards the DSLLine Info attributes. dsl-line-info-forwarding
Example:
RP/0/RSP0/CPU0:router(config-vpdn-template)# dsl-line-info-forwarding
Step 6
Sets IP ToS value for tunneled traffic.
The service value ranges from 0 to
255.
ip tos type_of_service_value
Example:
RP/0/RSP0/CPU0:router(config-vpdn-template)# ip tos 56
Step 7
Specifies tunnel for a vpn and
configures the vpn id with the value
vpn id value
Example:
RP/0/RSP0/CPU0:router(config-vpdn-temp)# vpn id 3333:33
Step 8
3333:33. The value ranges from
0-ffffff in hexadecimal.
Release 4.3.x
106 OL-28375-03
Configures the vpn vrf name. vpn vrf vrf_name
Example:
RP/0/RSP0/CPU0:router(config-vpdn-template)# vpn vrf vrf_1
Step 9
commitSaves the configuration
changes, and remains within the
endPrompts user to take one of
these actions:
Yes Saves configuration
changes and exits the
NoExits the configuration
session without committing the
CancelRemains in the
configuration mode, without
committing the configuration
changes.
Configuring the VPDN Template: An example
configure
l2tp-class class hello-interval 100
vpdn
template l2tp-class class //template default will be used and display in show run
template tunnel busy timeout 567
l2tp-class class
vpdn
template default
l2tp-class class
!
end
Configuring Maximum Simultaneous VPDN Sessions
Perform this task to configure the maximum simultaneous vpdn sessions for session limiting per tunnel:
4.3.x
OL-28375-03 107
SUMMARY STEPS
1. configure
2. vpdn
3. session-limit number_of_sessions
DETAILED STEPS
Example:
Step 1
Enables VPDN and enters the VPDN sub-mode. vpdn
Example:
Step 2
Configures the maximumsimultaneous VPDNsessions. The
range is from 1 to 131072.
session-limit number_of_sessions
Example:
RP/0/RSP0/CPU0:router(config-vpdn)# session-limit
200
Step 3
If limit is configured after a number of sessions are
up, then those sessions remain up irrespective of the
limit.
Note
Configuring Maximum Simultaneous VPDN Sessions: An example
configure
vpdn
session-limit 200
!
end
Release 4.3.x
108 OL-28375-03
Activating VPDN Logging
Perform this task to activate logging of VPDN event information. When VPDN event logging is enabled,
VPDN event messages are logged as the events occur.
Tunnel start and stop records are generated without any tunnel statistics. Note
SUMMARY STEPS
1. configure
2. vpdn
3. logging [cause| cause-normal | dead-cache | local | tunnel-drop | user ]
4. history failure
DETAILED STEPS
Example:
Step 1
Enters the VPDN sub-mode. vpdn
Example:
Step 2
Enables the logging of generic VPDN events. logging [cause| cause-normal | dead-cache | local |
tunnel-drop | user ]
Step 3
Example:
RP/0/RSP0/CPU0:router(config-vpdn)# logging
local
user
cause
tunnel-drop
Enables logging of VPDN failure events to the history failure
table.
history failure
Example:
RP/0/RSP0/CPU0:router(config-vpdn)# history
failure
Step 4
4.3.x
OL-28375-03 109
Activating VPDN Logging: An example
configure
vpdn
history failure
logging local
logging user
logging cause-normal
logging tunnel-drop
logging dead-cache
!
end
Configuring Options to Apply on Calling Station ID
Perform this task to configure options to apply on calling station ID. The calling station ID provides detailed
information about the originator of the session, such as the phone number of the originator, the Logical Line
ID(LLID) used to make the connection on the LAC, or the MACaddress of the PCconnecting to the network.
SUMMARY STEPS
1. configure
2. vpdn
3. caller-id mask-method remove match match_name
Release 4.3.x
110 OL-28375-03
DETAILED STEPS
Example:
Step 1
Example:
Step 2
Suppresses the calling station IDfor all users. If there is a match
option, then calling station ID only for users which have the
match-string in their username is suppressed.
caller-id mask-method remove match match_name
Example:
RP/0/RSP0/CPU0:router(config-vpdn)# caller-id
mask-method remove match match_class
Step 3
This command an also be run under the vpdn template
configuration mode.
Note
Configuring Options to Apply on Calling Station ID: An example
configure
vpdn //or vpdn template
caller-id mask-method remove match match_call
!
end
Configuring L2TP Session-ID Commands
Perform this task to configure L2TP session-id commands.
4.3.x
OL-28375-03 111
SUMMARY STEPS
1. configure
2. vpdn
3. l2tp session-id space hierarchical
DETAILED STEPS
Example:
Step 1
Configures vpdn. vpdn
Example:
Step 2
Enables the hierarchical session-ID allocation algorithm. l2tp session-id space hierarchical
Example:
session-id space hierarchical
Step 3
session.
Configuring L2TP Session-ID Commands: An example
configure
vpdn
l2tp session-id space hierarchical
!
end
Release 4.3.x
112 OL-28375-03
Configuring L2TP Class Options
Perform this task to configure the various options for L2TP class.
SUMMARY STEPS
1. configure
2. l2tp-class class_name
3. authentication [ disable | enable ]
4. congestion control
5. digest [check disable |hash { MD5 | SHA1 } | secret { 0 | 7 | LINE } ]
6. hello-interval interval_duration
7. hostname host_name
8. receive-window size
9. retransmit initial [retries | retries_number | timeout {max max_seconds | min min_seconds }
10. timeout [no-user{ timeout_value | never } | setup setup_value ]
11. tunnel accounting accounting_method_list_name
DETAILED STEPS
Example:
Step 1
Configures the L2TP class command. l2tp-class class_name
Example:
RP/0/RSP0/CPU0:router(config)# l2tp-class class1
Step 2
Enables the tunnel authentication. The Enable and
Disable options enables or disables the L2TP tunnel
authentication.
authentication [ disable | enable ]
Example:
RP/0/RSP0/CPU0:router(config-l2tp-class)# authentication
disable
Step 3
Enables L2TP congestion control. congestion control
Example:
RP/0/RSP0/CPU0:router(config-l2tp-class)# congestion
control
Step 4
4.3.x
OL-28375-03 113
Messages the Digest configuration for L2TPv3
control connection.
digest [check disable |hash { MD5 | SHA1 } | secret { 0 | 7 | LINE
} ]
Example:
RP/0/RSP0/CPU0:router(config-l2tp-class)# digest check
Step 5
disable
RP/0/RSP0/CPU0:router(config-l2tp-class)# digest hash
MD5
RP/0/RSP0/CPU0:router(config-l2tp-class)# digest secret
0
Sets HELLOmessage interval for specified amount
of seconds.
hello-interval interval_duration
Example:
RP/0/RSP0/CPU0:router(config-l2tp-class)# hello-interval
45
Step 6
Sets the local hostname for control connection
authentication.
hostname host_name
Example:
RP/0/RSP0/CPU0:router(config-l2tp-class)# hostname
local_host
Step 7
Receives window size for the control connection.
The range is from 1 to 16384.
receive-window size
Example:
RP/0/RSP0/CPU0:router(config-l2tp-class)# receive-window
56
Step 8
The range is from 1 to 16384.
retransmit initial [retries | retries_number | timeout {max
max_seconds | min min_seconds }
Example:
RP/0/RSP0/CPU0:router(config-l2tp-class)# retransmit
Step 9
initial retries 58
initial timeout max 6
The timeout value range, in seconds, is from 0 to
86400. The setup value range is from 60 to 6000.
timeout [no-user{ timeout_value | never } | setup setup_value ]
Example:
RP/0/RSP0/CPU0:router(config-l2tp-class)# timeout no-user
Step 10
56
setup 60
Configures the AAAaccounting method list name. tunnel accounting accounting_method_list_name
Example:
RP/0/RSP0/CPU0:router(config-l2tp-class)# tunnel
Step 11
accounting acc_tunn
Release 4.3.x
114 OL-28375-03
changes.
Configuring L2TP Class Options: An example
configure
l2tp-class class1
authentication enable
congestion-control
digest check disable
hello-interval 876
hostname l2tp_host
receive-window 163
retransmit initial timeout 60
timeout no-user 864
tunnel accounting aaa_l2tp
!
end
Configuring Softshut for VPDN
Perform this task to configure softshut for vpdn.
SUMMARY STEPS
1. configure
2. vpdn
3. softshut
4.3.x
OL-28375-03 115
DETAILED STEPS
Example:
Step 1
Example:
Step 2
Ensures that no new sessions are allowed. softshut
Example:
RP/0/RSP0/CPU0:router(config-vpdn)# softshut
Step 3
Configuring Softshut for VPDN: An example
configure
vpdn
softshut
!
end
PPPoE Smart Server Selection
The PPPoE Smart Server Selection (PADO delay) feature in BNG allows the PPPoE client to control the
selection of BNG for session establishment, in a multi-BNG setup. The feature provides the option for
configuring a delay in sending PADO messages from BNG, in response to the PADI messages received from
the PPPoE clients. This, in turn, helps in establishing a priority order and load balancing across all BNGs.
When establishing a PPPoE session in a multi-BNG setup, the clients broadcast their PADI messages to all
BNGs. When the BNGs reply with a PADO message, the subscriber selects a BNG, and sends a PADR
Release 4.3.x
116 OL-28375-03
message to the BNG with which a session needs to be established. Most PPPoE clients send a PADR message
to the BNGfromwhich it received the first PADOmessage. By configuring the Smart Server Selection feature
on BNG, a delay is added to the PADO messages sent from the BNG, based on the properties of the PADI
messages received fromthe PPPoE clients. This delay in receiving the PADOpackets, in turn, gives the PPPoE
client the flexibility of effectively selecting the appropriate BNG to which the PADR message is to be sent.
Configuration options for Smart Server Selection
Allows configuring a specific delay for the PADO message sent from BNG.
Allows configuring a delay for the PADO message sent from BNG, based on the Circuit-ID, Remote-ID
and Service-Name contained in the incoming PADI message.
Allows Circuit-ID and Remote-ID tag matching, with strings up to 64 characters in length.
Allows partial matching on Circuit-ID, Remote-ID, and Service-Name contained in the incoming PADI
message.
For configuring the delay for a PADO message, see Configuring PADO Delay, on page 117.
Configuring PADO Delay
Perform this task to configure a delay for PPPoE Active Discovery Offer (PADO) message, or in other words,
enabling Smart Server Selection feature for a PPPoE BBA-Group in BNG.
If multiple delays match a particular subscriber, Circuit-ID matches are preferred to Remote-ID matches,
which, in turn, are preferred to Service-Name matches.
Note
SUMMARY STEPS
1. configure
2. pppoe bba-group bba-group-name
3. Use these commands to configure the PADOdelay based on a specific delay value, Circuit-ID, Remote-ID,
and Service-Name respectively:
pado delay delay
pado delay circuit-id {delay | {string | contains} string delay}
pado delay remote-id {delay | {string | contains} string delay}
pado delay service-name {string | contains} string delay
4.3.x
OL-28375-03 117
DETAILED STEPS
Example:
Step 1
Enters the PPPoE BBA-Group configuration mode. pppoe bba-group bba-group-name
Example:
bba_1
Step 2
Sets the PADO delay in milliseconds based on: Use these commands to configure the PADOdelay based
on a specific delay value, Circuit-ID, Remote-ID, and
Service-Name respectively:
Step 3
A specific delay value
Circuit-ID received in PADI
pado delay delay
Remote-ID received in PADI
pado delay circuit-id {delay | {string | contains}
string delay}
Service-Name received in PADI
pado delay remote-id {delay | {string | contains}
string delay}
The delay range is from 0 to 10000.
The string option delays the PADOmessage, when the Circuit-ID
(or Remote-ID or Service-Name) received in the PADI message
matches the configured string value.
pado delay service-name {string | contains}
string delay
The contains option delays the PADO message, when the
Circuit-ID(or Remote-IDor Service-Name) received in the PADI
message contains the configured string value.
Example:
RP/0/RSP0/CPU0:router(config-bbagroup)# pado
delay 500
delay circuit-id 200
delay remote-id string circuit4
delay service-name contains service 9950
Release 4.3.x
118 OL-28375-03
Configuring PPPoE PADO delay : An example
pppoe bba-group bba_1
pado delay 500
pado delay remote-id 100
pado delay circuit-id string circuit4 8000
pado delay service-name contains service 9950
!
end
PPPoE Session Limit and Throttle
PPPoE Session Limit
The PPPoE Session Limit support limits the number of PPPoE sessions that can be created on a BNG router.
As a result, it reduces excessive memory usage by the BNG router for virtual access.
This offers additional configuration flexibility on the BNG router by limiting the number of PPPoE sessions
for each:
Line card
Parent interface
Peer MAC address
Peer MAC address under individual access interface
Circuit-ID
Remote-ID
Combination of Circuit-ID and Remote ID
Access interface using the same Inner VLAN tag
Access interface using the same Outer VLAN tag.
Access interface using the same Inner and Outer VLAN tags
The PPPoE Session Limit support also limits the number of Inter Working Function (IWF) sessions for each
peer MAC address and for each peer MAC address under individual access interface.
See, Configuring PPPoE Session Limit, on page 119.
Configuring PPPoE Session Limit
Perform this task to configure PPPoE session limit for a PPPoE BBA-Group in BNG.
SUMMARY STEPS
1. configure
3. sessions {access-interface | circuit-id | circuit-id-and-remote-id | inner-vlan | {{mac | mac-iwf}
[access-interface] }} | max | outer-vlan | remote-id | vlan} limit limit-count [threshold threshold-count]
4.3.x
OL-28375-03 119
DETAILED STEPS
Example:
Step 1
Enters the PPPoE BBA-Group configuration mode. pppoe bba-group bba-group name
Example:
bba_1
Step 2
Configures the PPPoE session limits. sessions {access-interface | circuit-id |
circuit-id-and-remote-id | inner-vlan | {{mac |
Step 3
If the optional argument, threshold is configured, a log message
is generated when the PPPoE session limit exceeds the
threshold-count value.
mac-iwf} [access-interface] }} | max | outer-vlan |
remote-id | vlan} limit limit-count [threshold
threshold-count]
The limit-count value and threshold-count value ranges from 1
to 65535. The default value is 65535.
Example:
RP/0/RSP0/CPU0:router(config-bbagroup)# sessions
access-interface limit 1000
mac access-interface limit 5000 threshold 4900
circuit-id limit 8000 threshold 7500
Configuring PPPoE Session Limit: An example
configure
pppoe bba-group bba1
sessions circuit-id limit 8000 threshold 7500
sessions access-interface limit 1000
sessions mac access-interface limit 5000 threshold 900
!
end
Release 4.3.x
120 OL-28375-03
PPPoE Session Throttle
The PPPoE Session Throttle support on BNG limits the number of PPPoE session requests coming to BNG
within a specified period of time. This, in turn, ensures that the session establishment of other client requests
coming to the BNG server is not impacted.
This offers configuration flexibility in the BNG router by throttling the number of session requests based on
one of these:
Peer MAC address
Peer MAC address under individual access interface
Circuit-ID
Remote-ID
A combination of Circuit-ID and Remote ID
Inner VLAN tag under individual access interface
Outer VLAN tag under individual access interface
Inner and Outer VLAN tag under individual access interface
The PPPoE session throttle support also throttles the number of Inter Working Function (IWF) session requests
for each peer MAC address under an individual access interface.
See, Configuring PPPoE Session Throttle, on page 121.
Configuring PPPoE Session Throttle
Perform this task to configure PPPoE session throttle for a PPPoE BBA-Group in BNG.
SUMMARY STEPS
1. configure
3. sessions {circuit-id | circuit-id-and-remote-id | inner-vlan | {mac [access-interface] } | {mac-iwf
{access-interface}} | outer-vlan | remote-id | vlan} throttle request-count request-period blocking-period
DETAILED STEPS
Example:
Step 1
4.3.x
OL-28375-03 121
Enters the PPPoE BBA-Group configuration mode. pppoe bba-group bba-group name
Example:
bba_1
Step 2
Configures the PPPoE session throttles. sessions {circuit-id | circuit-id-and-remote-id |
inner-vlan | {mac [access-interface] } | {mac-iwf
Step 3
The request-count value ranges from 1 to 65535.
{access-interface}} | outer-vlan | remote-id | vlan}
throttle request-count request-period blocking-period
The request-period value ranges from 1 to 100.
Example:
The blocking-period value ranges from 1 to 100.
circuit-id throttle 1000 50 25
mac-iwf access-interface throttle 5000 100 50
Configuring PPPoE Session Throttle: An example
configure
pppoe bba-group bba1
sessions circuit-id throttle 1000 50 25
sessions mac-iwf access-interface throttle 5000 100 50
!
Making DHCP Settings
The Dynamic Host Configuration Protocol (DHCP) is a network protocol used to configure network devices
so that they can communicate on an IP network. There are three distinct elements in a DHCP network:
DHCP clientIt is the device that seeks IP configuration information, such as IP address.
DHCP serverIt allocates IP address from its address pool to the DHCP client.
Release 4.3.x
122 OL-28375-03
Making DHCP Settings
DHCP relay or DHCP proxyIt passes IP configuration information between the client and server. It
is used when DHCP client and DHCP server are present on different networks.
Initially, the DHCP client (which is a CPE) does not possess an IP address. As a result, it sends a L2 broadcast
request to get an IP address. Acting as the relay agent, BNGprocesses the request and forwards it to the DHCP
server. BNG also forwards responses from the DHCP server back to the DHCP client, ensuring that the end
device gets correct IP configuration information. A typical DHCP layout is depicted in this figure.
Figure 8: DHCP Network
The DHCP server allocates IP addresses for only a configurable period of time known as the lease period. If
a client device needs to retain the IP address for a period longer than the lease period, then the client must
renew the lease before it expires. To renew the lease, the client sends a unicast request to the DHCP server.
On receiving the request message, the server responds with an acknowledgment, and the client's lease is
extended by the lease time specified in the acknowledgment message.
When a control policy is applied to an access interface, it becomes a subscriber access interface. Otherwise,
it is a DHCP standalone interface. For the standalone interface, DHCP adds routes to RIB and populates ARP
entries, based on the configuration.
For the subscriber access interface, DHCP uses the policy-plane to determine whether the IP subscriber session
should be created for a client binding. This is determined based on whether a valid control policy is applied
to the access-interface on which the client binding is created. If a subscriber session is created, then a route
is added for the subscriber interface, but no ARP requests are sent out from that subscriber interface.
BNG can be configured to either act as DHCP proxy in the DHCP network.
DHCP relay is not supported for BNG. Note
Enabling DHCP Proxy
As the DHCP proxy, BNG performs all the functions of a relay and also provides some additional functions.
In the proxy mode, BNG conceals DHCP server details from DHCP clients. BNG modifies the DHCP replies
4.3.x
OL-28375-03 123
Enabling DHCP Proxy
such that the client considers the proxy to be the server. In this state the client interacts with BNG as if it is
the DHCP server.
BNG procures IP leases from the DHCP server and keeps it in its pool. When the client needs to renew its
lease, it unicasts the lease renewal request directly to the BNG, assuming it to be the server. BNG renews the
lease by allocating the lease from its lease pool.
This way the DHCP proxy splits the lease management process into two phases:
Server to Proxy (Proxy Lease)
Proxy to Client (Client lease)
The two phase lease management has these features:
Shorter client lease times and longer proxy lease times.
High frequency lease management (renews) at network edge.
Low frequency lease management (renews) at centralized server.
The benefits of DHCP proxy are:
Reduced traffic between BNG and DHCP server.
Quicker client response to network outages.
Configuring DHCP proxy on BNG involves these phases:
Creating a proxy profile. The profile contains various proxy settings. These settings are applied when
the profile is attached to an interface. To create a proxy profile, see Configuring DHCP IPv4 Profile
Proxy Class, on page 124
Specifying client lease period. The client should renewthe lease before the completion of this time
period, otherwise the lease expires. To specify the client lease period within a proxy profile, see
Configuring the Client Lease Time, on page 128.
Specifying remote-ID. The remote-ID is used by the proxy to identify the host that had sent the
DHCP request. To define a remote-id within a proxy profile, see Configuring a Remote-ID, on
page 127.
Specifying circuit-IDfor an interface. The circuit-IDis used by the proxy to identify the circuit in which
the DHCP request was received. Later, DHCP proxy uses it for relaying DHCP responses back to the
proper circuit. The circuit-ID is defined for an interface. To define it, see Configuring a Circuit-ID for
an Interface, on page 126.
Attaching proxy profile to an interface. See, Attaching a Proxy Profile to an Interface, on page 130
Configuring DHCP IPv4 Profile Proxy Class
Perform this task to define DHCP.
Release 4.3.x
124 OL-28375-03
Enabling DHCP Proxy
SUMMARY STEPS
1. configure
2. dhcp ipv4
3. profile profile-name proxy
4. class class-name
6. show dhcp ipv4 proxy profile name name
DETAILED STEPS
Example:
Step 1
Enters the IPv4 DHCP configuration mode. dhcp ipv4
Example:
RP/0/RSP0/CPU0:router(config)# dhcp ipv4
Step 2
Enters the proxy profile configuration mode. The DHCP Proxy
makes use of the class information to select a subset of
parameters in a given profile.
profile profile-name proxy
Example:
RP/0/RSP0/CPU0:router(config-dhcpv4)# profile
profile1 proxy
Step 3
Creates a DHCP proxy profile class and enters the proxy profile
class mode.
class class-name
Example:
RP/0/RSP0/CPU0:router(config-dhcpv4-profile)#
class blue
Step 4
4.3.x
OL-28375-03 125
Enabling DHCP Proxy
(Optional) Displays the details proxy profile information. show dhcp ipv4 proxy profile name name
Example:
RP/0/RSP0/CPU0:routershow dhcp ipv4 proxy
profile name profile1
Step 6
Configuring a Circuit-ID for an Interface
Perform this task to configure a circuit-id for an interface.
SUMMARY STEPS
1. configure
2. dhcp ipv4
4. proxy information option format-type circuit-id value
DETAILED STEPS
Example:
Step 1
Enters DHCP IPv4 configuration submode. dhcp ipv4
Example:
Step 2
Configures the interface and enters the interface configuration
mode.
Example:
RP/0/RSP0/CPU0:router(config-dhcpv4)#
interface Bundle-Ether 355
Step 3
Configures the circuit-id for this interface. proxy information option format-type circuit-id
value
Step 4
Example:
RP/0/RSP0/CPU0:router(config-dhcpv4)# proxy
information option format-type circuit-id
7
Release 4.3.x
126 OL-28375-03
Enabling DHCP Proxy
Configuring a Circuit-ID for an Interface: An example
configure
dhcp ipv4
proxy information option format-type circuit-id 7
!
!
end
Configuring a Remote-ID
Perform this task to configure a remote-ID.
SUMMARY STEPS
1. configure
2. dhcp ipv4
4. relay information option remote-id value
DETAILED STEPS
Example:
Step 1
4.3.x
OL-28375-03 127
Enabling DHCP Proxy
Example:
Step 2
Creates a DHCP proxy profile. profile profile-name proxy
Example:
profile profile1 proxy
Step 3
Inserts relay agent information for remote id suboptions such as
remote-ID value.
relay information option remote-id value
Example:
RP/0/RSP0/CPU0:router(config-if)# relay
information option remote-id 9
Step 4
Configuring a Remote-ID: An example
configure
dhcp ipv4
relay information option remote-id 9
!
!
end
Configuring the Client Lease Time
Perform this task to configure the client lease time. It defines the time period after which the client lease
expires.
Release 4.3.x
128 OL-28375-03
Enabling DHCP Proxy
SUMMARY STEPS
1. configure
2. dhcp ipv4
4. lease proxy client-lease-time value
DETAILED STEPS
Example:
Step 1
Example:
Step 2
Creates a DHCP profile. profile profile-name proxy
Example:
profile1 proxy
Step 3
Configures a client lease time for each profile. The
minimum value of the lease proxy client time is 300
seconds.
lease proxy client-lease-time value
Example:
RP/0/RSP0/CPU0:router(config-dhcpv4-proxy-profile)#
lease proxy client-lease-time 600
Step 4
4.3.x
OL-28375-03 129
Enabling DHCP Proxy
Configuring the Client Lease Time: An example
configure
dhcp ipv4
lease proxy client-lease-time 600
!
!
end
Attaching a Proxy Profile to an Interface
Preform this task to attach a proxy profile to an interface. After it is attached, the various settings specified
in the proxy profile take effect on the interface.
SUMMARY STEPS
1. configure
2. dhcp ipv4
3. interface type interface-path-id proxy profile profile-name
5. show dhcp ipv4 proxy profile name name
DETAILED STEPS
Example:
Step 1
Example:
Step 2
Enters the Interface configuration mode and assigns a proxy
profile to an interface.
interface type interface-path-id proxy profile
profile-name
Example:
Step 3
interface Bundle-Ether 344 proxy profile
profile1
Release 4.3.x
130 OL-28375-03
Enabling DHCP Proxy
(Optional) Displays the details proxy profile information. show dhcp ipv4 proxy profile name name
Example:
RP/0/RSP0/CPU0:router# show dhcp ipv4 proxy
profile name profile1
Step 5
Attaching a Proxy Profile to an Interface: An example
configure
dhcp ipv4
interface Bundle-Ether100.10 proxy profile profile1
proxy information option format-type circuit-id 7
!
!
end
Specifying DHCP Lease Limit
The DHCP lease limit feature allows you to limit the number of DHCP bindings on an interface. A binding
represents the mapping between the MAC address of the client and the IP address allocated to it. The lease
limit can be specified for each Circuit-ID, or Remote-ID, or interface.
The lease limit can be configured through a DHCP proxy profile. When this profile is attached to an interface,
bindings up to the configured limit on that interface are allowed. For example, if a profile with a per-circuit
lease limit of 10 bindings is assigned to four interfaces, then for each unique Circuit-ID, there would be 10
bindings allowed for each interface.
If the lease limit is lowered below the current number of existing bindings, then the existing bindings are
allowed to persist, but no new bindings are allowed to be created until the number of bindings drops below
the new lease limit.
If the lease limit is specified fromthe AAAserver, as part of Change of Authorization (CoA) or Access-Accept
message, then the DHCP lease limit configured through the proxy profile is overridden. In this case, the most
recent session limit, received from the AAA server, is taken as the current lease limit for the particular
Circuit-ID. The lease limit set fromthe AAAserver is cleared when there are no more client bindings associated
with the Circuit-ID for which the lease limit is applied.
To specify the lease limit, see these procedures:
Specifying the Lease Limit for a Circuit-ID, on page 132
Specifying the Lease Limit for a Remote-ID, on page 133
Specifying the Lease Limit for an Interface, on page 134
4.3.x
OL-28375-03 131
Specifying the Lease Limit for a Circuit-ID
Perform this task to specify the lease limit for each Circuit-ID.
SUMMARY STEPS
1. configure
2. dhcp ipv4
4. limit lease per-circuit-id value
DETAILED STEPS
Example:
Step 1
Example:
Step 2
Example:
profile1 proxy
Step 3
Specifies the lease limit for a Circuit-ID that is applied
to an interface.
limit lease per-circuit-id value
Example:
limit lease per-circuit-id 1000
Step 4
Release 4.3.x
132 OL-28375-03
Specifying the Lease Limit for a Circuit-ID: An example
configure
dhcp ipv4
limit lease per-circuit-id 1000
!
!
end
Specifying the Lease Limit for a Remote-ID
Perform this task to specify the lease limit for each Remote-ID.
SUMMARY STEPS
1. configure
2. dhcp ipv4
4. limit lease per-remote-id value
DETAILED STEPS
Example:
Step 1
Example:
Step 2
Example:
profile1 proxy
Step 3
Specifies the lease limit for a Remote-ID that is applied
to an interface.
limit lease per-remote-id value
Example:
limit lease per-remote-id 1340
Step 4
4.3.x
OL-28375-03 133
Specifying the Lease Limit for a Remote-ID: An example
configure
dhcp ipv4
limit lease per-remote-id 1340
!
!
end
Specifying the Lease Limit for an Interface
Perform this task to specify the lease limit for each interface.
SUMMARY STEPS
1. configure
2. dhcp ipv4
4. limit lease per-interface value
DETAILED STEPS
Example:
Step 1
Example:
Step 2
Release 4.3.x
134 OL-28375-03
Example:
profile1 proxy
Step 3
Specifies the lease limit for each interface. limit lease per-interface value
Example:
limit lease per-interface 2400
Step 4
Specifying the Lease Limit for an Interface: An example
configure
dhcp ipv4
limit lease per-interface 2400
!
!
end
Understanding DHCP Option-82
DHCP Option 82 allows the DHCP server to generate IP addresses based on the location of the client device.
This option defines these sub-options:
Agent Circuit ID Sub-optionThis sub-option is inserted by DSLAM and identifies the subscriber line
in the DSLAM.
Agent Remote ID Sub-optionThis sub-option is inserted by DSLAM or BNG in an l2-connected
topology. It is the client MAC address, but can be overridden. With the DHCP proxy or relay, the client
MAC address is lost by the time the packet gets to the DHCP server. This is a mechanism that preserves
the client MAC when the packet gets to the server.
4.3.x
OL-28375-03 135
Understanding DHCP Option-82
VPN identifier sub-optionThis sub-option is used by the relay agent to communicate the VPN for
every DHCP request that is sent to the DHCP server, and it is also used to forward any DHCP reply that
the DHCP server sends back to the relay agent.
Subnet Selection Sub-optionThis sub-option allows the separation of the subnet from the IP address
and is used to communicate with the relay agent. In a DHCP processing, the gateway address specifies
both the subnet on which a DHCP client resides, and the IP address that the server uses to communicate
with the relay agent.
Server Identifier Override Sub-optionThis sub-option value is copied in the reply packet from the
DHCP server, instead of the normal server ID address. This sub-option contains the incoming interface
IP address, which is the IP address on the relay agent that is accessible from the client. Using this
information, the DHCP client sends all renew and release packets to the relay agent, which in turn adds
all of the VPN sub-options and forwards the renew and release packets to the original DHCP server.
The VPN Identifier, Subnet Selection, and Server Identifier Override sub-options are used by DHCP
relay/proxy for supporting MPLS VPNs.
Note
Option 82 Relay Information Encapsulation
When two relay agents are relaying messages between the DHCP client and DHCP server, the second relay
agent (closer to the server), by default, replaces the first option 82 information with its own option 82. The
remote ID and circuit ID information from the first relay agent is lost. In some deployment scenarios, it is
necessary to maintain the initial option 82 from the first relay agent, in addition to the option 82 from the
second relay agent.
The DHCP option 82 relay information encapsulation feature allows the second relay agent to encapsulate
option 82 information in a received message from the first relay agent, if it is also configured to add its own
option 82 information. This configuration allows the DHCP server to use option 82 information from both
the relay agents.
Configuring DHCPv4 Class of Service (CoS)
BNG supports manual reset of Class of Service (CoS) value of DHCPv4 control packets sent on subscriber
interfaces. By default, the outer and inner CoS values are set to 6. This feature allows to set or modify these
CoS values sent by BNG.
To reset the CoS values, use the dhcp ipv4 [inner-cos | outer-cos] value command.
DHCPv6 Overview
The Dynamic Host Configuration Protocol for IPv6 (DHCPv6) enables DHCP servers to pass configuration
parameters, such as IPv6 network addresses, to IPv6 nodes. It enables automatic allocation of reusable network
addresses to the requesting clients, using the stateful address-configuration. Along with address and prefix
allocation, DHCPv6 also offers additional configuration flexibility by assigning other configuration parameters
such as DNS address, DNS domain name, AFTR address to IPv6 nodes in a network.
Release 4.3.x
136 OL-28375-03
Configuring DHCPv4 Class of Service (CoS)
The basic DHCPv6 client-server concept is similar to using DHCP for IPv4 (DHCPv4). If a client wishes to
receive configuration parameters, it sends out a request on the attached local network to detect the available
DHCPv6 servers. Although DHCPv6 assigns IPv6 addresses or prefixes, name servers, and other configuration
information very similar to that of DHCP for IPv4, these are certain key differences between DHCPv4 and
DHCPv6. For example, unlike DHCPv4, address allocation in DHCPv6 is handled using a message option,
DHCPv6 clients can request multiple addresses and prefixes in a single request, and DHCPv6 can request
different lease times for the addresses and prefixes. These significant advantages of DHCPv6 make it a
preferred protocol for address assignment.
IPv6 hosts use Stateless Address Auto-Configuration (SLAAC), a model in which the hosts generate their
own addresses using a combination of local and router-advertised information.
The DHCPv6 has been standardized by the IETF through RFC 3315. This DHCPv6 protocol is a stateful
counterpart to IPv6 Stateless Address Auto-Configuration (RFC 4862), and can be used separately, or
concurrently with SLAAC, to obtain configuration parameters.
Prior to configuring DHCPv6, IPv6 must be enabled on the interface on which DHCPv6 is servicing and
enable Neighbor Discovery (ND).
For more information about Neighbor Discovery (ND), refer to the "Implementing Network Stack IPv4
and IPv6" section in the Cisco IOS XR IP Addresses and Services Configuration Guide.
Note
Restrictions
DHCPv6 Proxy supports to a maximum of eight external DHCPv6 servers per proxy profile.
Bulk lease query is not supported.
DHCPv6 server is supported only with BNG configuration.
DHCPv6 Server and DHCPv6 Proxy
The DHCPv6 server always uses stateful address assignment. On receiving a valid request, the DHCPv6
server assigns IPv6 address or prefix and other configuration attributes such as domain name, domain name
server (DNS) address to requesting clients.
A DHCPv6 Relay or Proxy forwards a DHCPv6 message from a client to a server. A DHCPv6 Relay can use
either stateless or stateful address assignment. The DHCPv6 Stateless Relay agent acts as an intermediary to
deliver DHCPv6 messages between clients and servers. The Relay does not store or keep track of information
such as client addresses or the lease time. The DHCPv6 Relay is also known as a Stateless Relay. On the other
hand, the DHCPv6 Stateful Relay agent, also known as DHCP proxy, not only forwards a DHCPv6 message
froma client to the server, but also keeps track of the client's addresses and lease time. Hence, DHCPv6 Proxy
is also known as Stateful Relay. DHCPv6 supports a standalone proxy.
DHCPv6 Proxy enables inserting remote-IDand interface-IDoptions. The DHCPv6 Proxy uses the interface-ID
in addition to remote-ID to choose the interface on which to send the response towards client.
DHCPv6 can be enabled on different configuration modes. For more information about configuring DHCPv6
on different configuring modes, see Enabling DHCPv6 for Different Configuration Modes, on page 138. For
more information about setting the DHCPv6 parameters, see Setting Up DHCPv6 Parameters, on page 141.
4.3.x
OL-28375-03 137
DHCP relay is not supported for BNG. Note
Enabling DHCPv6 for Different Configuration Modes
Perform this task to enable DHCPv6 for different configuration modes such as global, server profile, proxy
profile configuration modes, and server profile class and proxy profile class sub-configuration modes.
SUMMARY STEPS
1. configure
2. dhcp ipv6
3. profile server_profile_name server
4. class class-name
5. dns-server address
6. domain-name name
7. prefix-pool pool_name
8. address-pool pool_name
10. interface type interface-path-id server profile profile_name
11. profile proxy_profile_name proxy
12. link-address ipv6_address
13. class class-name
14. helper-address vrf vrf_name ipv6_address
16. interface type interface-path-id proxy profile profile_name
DETAILED STEPS
Example:
Step 1
Configures DHCP for IPv6 and enters the DHCPv6
configuration mode.
dhcp ipv6
Example:
Step 2
Release 4.3.x
138 OL-28375-03
Creates a DHCPv6 server profile and enters the DHCPv6
server profile sub-configuration mode.
profile server_profile_name server
Example:
my-server-profile server
Step 3
Defines a class in a server profile and enters the server
profile class sub-mode.
class class-name
Example:
RP/0/RSP0/CPU0:router(config-dhcpv6-server-profile)#
class server-green
Step 4
Defines a dns-server and the corresponding address in
a server profile.
dns-server address
Example:
dns-server 1111::1
Step 5
Defines a domain name in a server profile. domain-name name
Example:
domain-name www.xyz.com
Step 6
Confiures a prefix pool in a server profile. prefix-pool pool_name
Example:
prefix_pool p1
Step 7
Confiures an address pool in a server profile. address-pool pool_name
Example:
address_pool p1
Step 8
Associates a DHCPv6 server configuration profile with
an IPv6 interface.
interface type interface-path-id server profile profile_name
Example:
RP/0/RSP0/CPU0:router(config-dhcpv6)# interface
Bundle-Ether1.1 server profile my-server-profile
Step 10
4.3.x
OL-28375-03 139
Creates a DHCPv6 profile proxy and enters the DHCPv6
proxy sub-configuration mode.
profile proxy_profile_name proxy
Example:
my-proxy-profile proxy
Step 11
Specifies the IPv6 address to be filled in the link-address
field of the Relay Forward message.
link-address ipv6_address
Example:
RP/0/RSP0/CPU0:router(config-dhcpv6)# link-address
5:6::78
Step 12
Defines a class in a proxy profile and enters the proxy
profile class sub-mode.
class class-name
Example:
class proxy-red
Step 13
Configures DHCPv6 address as a helper address to the
proxy.
helper-address vrf vrf_name ipv6_address
Example:
helper-address vrf my-server-vrf 1:1:1::1
Step 14
The helper address can be configured only
under the proxy profile and proxy profile class
sub-modes.
Note
Associates a DHCPv6 proxy configuration profile to an
IPv6 interface.
interface type interface-path-id proxy profile profile_name
Example:
BundleEther100.1 proxy profile my-proxy-profile
Step 16
Release 4.3.x
140 OL-28375-03
Enabling DHCPv6 for Different Configuration Modes: An example
configure
dhcp ipv6
profile my-server-profile server
link-address 5:6::78
class server-green
dns-server 1111::1
domain-name www.cisco.com
prefix-pool POOL_P6_2
address-pool POOL_A6_1
end
!!
configure
dhcp ipv6
interface GigabitEthernet 0/2/0/0 server profile my-server-profile
profile my-proxy-profile proxy
link-address 5:6::78
class proxy-red
helper-address 5661:11
end
!!
configure
dhcp ipv6
interface GigabitEthernet 0/2/0/0 proxy profile my-proxy-profile
end
!!
Setting Up DHCPv6 Parameters
Perform this task to set up DHCPv6 parameters such as address pool name, prefix pool name, DNS server,
domain name, lease time, and helper address.
SUMMARY STEPS
1. configure
2. dhcp ipv6
4. dns-server ipv6_address
5. domain-name domain_name
6. lease
8. prefix-pool prefix-pool-name
9. address-pool address-pool-name
4.3.x
OL-28375-03 141
DETAILED STEPS
Example:
Step 1
configuration mode.
dhcp ipv6
Example:
Step 2
Configures DHCPv6 server profile and enters the
DHCPv6 server profile sub-configuration mode.
Example:
Step 3
Configures the DNS server for DHCPv6 server
profile.
dns-server ipv6_address
Example:
dns-server 1:1:1::1
Step 4
The DNS server name is defined in the
class mode. If the same parameters are
defined in the profile mode too, then the
values defined in the class mode takes
precedence.
Note
Configures the DNS domain name for DHCPv6
server profile.
domain-name domain_name
Example:
domain-name my.domain.name
Step 5
The DNS server name is defined in the
class mode. If the same parameters are
defined in the profile mode too, then the
values defined in the class mode takes
precedence.
Note
Configures the lease time for a duration of 1 day,
6 hours, and 0 minutes.
lease
Example:
RP/0/RSP0/CPU0:router(config-dhcpv6-server-profile)# lease
1 6 0
Step 6
Configures DHCPv6 address as a helper address
to the proxy.
Example:
Step 7
under the proxy profile and proxy profile
class sub-modes.
Note
Configures the prefix pool under the DHCPv6
server profile class sub-mode.
prefix-pool prefix-pool-name
Example:
RP/0/RSP0/CPU0:router(config-dhcpv6-server-profile-class)#
prefix-pool my-server-delegated-prefix-pool
Step 8
Release 4.3.x
142 OL-28375-03
Configures the address pool under the DHCPv6
server profile class sub-mode.
address-pool address-pool-name
Example:
RP/0/RSP0/CPU0:router(config-dhcpv6-server-profile-class)#
address-pool my-server-address-pool
Step 9
changes.
Setting Up DHCPv6 Parameters: An example
configure
dhcp ipv6
dns-server 1:1:1::1
domain-name my.domain.name
lease 1 6 0
class class1
prefix-pool my-server-delegated-prefix-pool
address-pool my-server-address-pool
end
!!
DHCPv6 Features
DHCPv6 is widely used in LAN environments to dynamically assign host IP addresses from a centralized
server. This dynamic assignment of addresses reduces the overhead of administration of IP addresses. DHCPv6
also helps conserve the limited IP address space. This is because IP addresses no longer need to be permanently
assigned to hosts; only those hosts that are connected to the network consume IP addresses.
The DHCPv6 features supported in BNG are:
High Availability Support for DHCPv6
High availability support for DHCPv6 includes:
4.3.x
OL-28375-03 143
DHCPv6 Features
Linecard Online Insertion and Removal
Linecard Online Insertion and Removal (OIR) enables you to replace faulty parts without affecting the system's
operations. When a card is inserted, power is available on the card, and it initializes itself to start being
operational.
DHCPv6 bindings are not affected by Linecard OIR. Note
Checkpoint and Shadow Database
The checkpoint and shadow database are actively maintained on the RSP and contains a copy of all bindings
from all linecards. The checkpoint database has client or subscriber bindings from the subscribers over
interfaces in its scope. The shadow database on the active RSP updates the standby shadow database.
DHCPv6 Hot Standby
DHCPv6 Hot Standby is a process that is supported only on RSPs. Whenever the active RSP stops responding,
it is instantly replaced by a standby RSP. The standby RSP takes over processing when it becomes active.
DHCPv6 Prefix Delegation
The DHCPv6 prefix delegation is a mechanism of delegating IPv6 prefixes to a client. The prefix delegation
feature can be used to manage link, subnet, and site addressing changes.
An Internet Service Provider (ISP) assigns prefix to a customer for use within the customer's network. Prefix
delegation occurs between a provider edge (PE) device and customer premises equipment (CPE), using the
DHCPv6 prefix delegation option. After the ISP has delegated prefixes to a customer, the customer may
further subnet and assign prefixes to the links in the customer's network.
By default, the prefix delegation feature is always enabled.
IPv6 IPoE Subscriber Support
An IPv6 subscriber transmits IPv6 address that is created using the DHCPv6 protocol. The IPv6 subscribers
run IPv6 on the CPE device and are connected to BNG through a Layer-2 network or through Layer-2
aggregation. The IPv6 subscribers are supported when they are directly connected to the BNG or through a
Layer-2 aggregator.
To enable IPv6 IPoE subscriber support, the DHCPv6 profile needs to be explicitly configured on the subscriber
interface. For more information, see Configuring IPv6 IPoE Subscriber Interface, on page 144.
FSOL Handling
The DHCPv6 First Sign of Life (FSOL) handling is only supported for IPoE sessions. DHCPv6 handles
SOLICIT packet from client as FSOL packet for IPoE session validation and creation. The IPoE session gets
created, as long as the configuration exists and the subscriber information is validated successfully.
Configuring IPv6 IPoE Subscriber Interface
Perform this task to configure IPoE subscriber interface.
Release 4.3.x
144 OL-28375-03
DHCPv6 Features
SUMMARY STEPS
1. configure
2. pool vrf name ipv6 pool_name
3. address-range first_ipv6_address last_ipv6_address
4. pool vrf name ipv6 pool_name
5. prefix-length length
6. prefix-range first_ipv6_address last_ipv6_address
8. dhcp ipv6
9. interface type interface-path-id server profile profile_name
11. prefix-pool pool_name
12. address-pool pool_name
14. dhcp ipv6
15. interface type interface-path-id proxy profile profile_name
16. profile server_profile_name proxy
19. dynamic-template type ipsubscriber dynamic_template_name
20. ipv6 enable
21. dhcpv6 address-pool pool_name
22. dhcpv6 delegated-prefix-pool pool_name
24. class-map type control subscriber match-all class-map_name
25. match protocol dhcpv6
26. end-class-map
27. policy-map type control subscriber class-map_name
29. class type control subscriber class_name do-all
31. end-policy-map
34. ipv4 address ipv4_address
35. ipv6 address ipv6_address
36. ipv6 enable
37. service-policy type control subscriber name
38. ipsubscriber ipv6 l2-connected
39. initiator dhcp
4.3.x
OL-28375-03 145
DHCPv6 Features
DETAILED STEPS
Example:
Step 1
Configures the distributed address pool service. pool vrf name ipv6 pool_name
Example:
RP/0/RSP0/CPU0:router(config)# pool vrf default ipv6
pool1
Step 2
Configures the address-range. address-range first_ipv6_address last_ipv6_address
Example:
RP/0/RSP0/CPU0:router(config-pool-ipv6)#
address-range 2201:abcd:1234:2400:f800::1
2201:abcd:1234:2400:f800::fff
Step 3
Configures the distributed address pool service. pool vrf name ipv6 pool_name
Example:
RP/0/RSP0/CPU0:router(config)# pool vrf default ipv6
pool2
Step 4
Specifies the prefix-length to be used. prefix-length length
Example:
RP/0/RSP0/CPU0:router(config-pool-ipv6)#
prefix-length 92
Step 5
Specifies the prefix-range for allocation. prefix-range first_ipv6_address last_ipv6_address
Example:
RP/0/RSP0/CPU0:router(config-pool-ipv6)# prefix-range
3301:1ab7:2345:1200:f800::
3301:1ab7:2345:1200:f800:fff0::
Step 6
Release 4.3.x
146 OL-28375-03
DHCPv6 Features
configuration mode.
dhcp ipv6
Example:
Step 8
Associates a DHCPv6 proxy configuration profile to
an IPv6 interface.
interface type interface-path-id server profile profile_name
Example:
Bundle-Ether1.1 server profile foo
Step 9
Creates a DHCPv6 server profile and enters the
Example:
RP/0/RSP0/CPU0:router(config-dhcpv6)# profile foo
server
Step 10
Configures a prefix pool in a server profile. prefix-pool pool_name
Example:
prefix-pool pool2
Step 11
Configures an address pool in the server profile. address-pool pool_name
Example:
address-pool pool1
Step 12
configuration mode.
dhcp ipv6
Example:
Step 14
Associates a DHCPv6 proxy configuration profile to
an IPv6 interface.
interface type interface-path-id proxy profile profile_name
Example:
Bundle-Ether1.1 proxy profile foo
Step 15
4.3.x
OL-28375-03 147
DHCPv6 Features
Creates a DHCPv6 server profile and enters the
profile server_profile_name proxy
Example:
RP/0/RSP0/CPU0:router(config-dhcpv6)# profile foo
proxy
Step 16
Configures DHCPv6 address as a helper address to the
proxy.
Example:
Step 17
under the proxy profile and proxy profile class
sub-modes.
Note
Configures the dynamic template of type ipsubscriber
and enters the dynamic template type configuration
mode.
dynamic-template type ipsubscriber dynamic_template_name
Example:
RP/0/RSP0/CPU0:router(config)# dynamic-template type
ipsubscriber dhcpv6_temp
Step 19
Enables IPv6 on an interface. ipv6 enable
Example:
ipv6 enable
Step 20
Configures DHCPv6 address pool. dhcpv6 address-pool pool_name
Example:
dhcpv6 address-pool pool3
Step 21
Configures DHCPv6 delegated prefix pool. dhcpv6 delegated-prefix-pool pool_name
Example:
dhcpv6 delegated-prefix-pool pool4
Step 22
Release 4.3.x
148 OL-28375-03
DHCPv6 Features
Configures the class map control subscriber with a
match-any criteria.
class-map type control subscriber match-all class-map_name
Example:
RP/0/RSP0/CPU0:router(config)# class-map type control
subscriber match-all dhcpv6_class
Step 24
Configures match criteria for the class configured in
the earlier step.
Example:
dhcpv6
Step 25
Configures the end class map. end-class-map
Example:
Step 26
Configures the subscriber control policy map. policy-map type control subscriber class-map_name
Example:
control subscriber dhcpv6-policy
Step 27
Configures the policy event with the match-first
criteria.
Example:
Step 28
match-any criteria.
class type control subscriber class_name do-all
Example:
control subscriber dhcpv6_class do-all
Step 29
Activates actions related to dynamic template. sequence_number activate dynamic-template
Step 30
Example:
dynamic-template dhcpv6_temp
Configures the end policy map. end-policy-map
Example:
RP/0/RSP0/CPU0:router(config-pmap-c)# end-policy-map
Step 31
4.3.x
OL-28375-03 149
DHCPv6 Features
Configures an interface and enters the interface
configuration mode.
Example:
Bundle-Ether1.1
Step 33
Configures the ipv4 address on an interface. ipv4 address ipv4_address
Example:
RP/0/RSP0/CPU0:router(config-if)# ipv4 address
11.11.11.2 255.255.255.0
Step 34
Configures the ipv6 address on an interface. ipv6 address ipv6_address
Example:
RP/0/RSP0/CPU0:router(config-if)# ipv6 address
11:11:11::2/64
Step 35
Example:
RP/0/RSP0/CPU0:router(config-if)# ipv6 enable
Step 36
interface.
service-policy type control subscriber name
Example:
RP/0/RSP0/CPU0:router(config-if)# service-policy type
control subscriber dhcpv6_policy
Step 37
Enables l2-connected IPv6 subscriber. ipsubscriber ipv6 l2-connected
Example:
RP/0/RSP0/CPU0:router(config-if)# ipsubscriber ipv6
l2-connected
Step 38
Configures IPv6 subscriber initiator. initiator dhcp
Example:
RP/0/RSP0/CPU0:router(config-if-ipsub-ipv6-l2conn)#
initiator dhcp
Step 39
Release 4.3.x
150 OL-28375-03
DHCPv6 Features
Configuring IPv6 IPoE Subscriber Interface: An example
configure
pool vrf default ipv6 pool1
address-range 2201:abcd:1234:2400:f800::1 2201:abcd:1234:2400:f800::fff
pool vrf default ipv6 pool2
prefix-length 92
prefix-range 3301:1ab7:2345:1200:f800:: 3301:1ab7:2345:1200:f800:fff0::
dhcp ipv6
interface GigabitEthernet0/3/0/0 server profile foo
profile foo server
prefix-pool pool2
address-pool pool1
!
!
end
configure
dhcp ipv6
interface GigabitEthernet0/3/0/0 proxy profile foo
profile foo proxy
helper address <v6 address of the server
!
!
dynamic-template type ipsubscriber dhcpv6_temp
ipv6 enable
dhcpv6 address-pool pool3
dhcpv6 delegated-prefix-pool pool4
!
!
!
class-map type control subscriber match-all dhcpv6_class
end-class-map
!
policy-map type control subscriber dhcpv6_policy
class type control subscriber dhcpv6_class do-all
20 activate dynamic-template dhcpv6_temp
!
!
end
configure
4.3.x
OL-28375-03 151
DHCPv6 Features
interface GigabitEthernet0/3/0/0
ipv4 address 11.11.11.2 255.255.255.0
ipv6 address 11:11:11::2/64
ipv6 enable
service-policy type control subscriber dhcpv6_policy
initiator dhcp
!
!
end
end
IPv6 PPPoE Subscriber Support
The PPPoE subscriber interfaces establish a PPP link with the subscriber, which is used for authentication
and address assignment. The DHCPv6 server assigns the address or prefix to the PPPoE subscriber. Because
the PPPoE subscriber interfaces are created dynamically, the DHCPv6 profile is applied to all the PPPoE
interfaces created on the router, and not just a single PPPoE interface.
To enable PPPoE subscriber support, you have to configure the DHCPv6 profile globally or on all PPPoE
interfaces. For more information, see Configuring IPv6 PPPoE Subscriber Interfaces, on page 152.
Configuring IPv6 PPPoE Subscriber Interfaces
Perform this task to configure PPPoE subscriber interfaces.
Release 4.3.x
152 OL-28375-03
DHCPv6 Features
SUMMARY STEPS
1. configure
2. dynamic-template type ppp dynamic_template_name
3. ppp authentication chap
4. ppp ipcp peer-address pool pool_name
5. ipv4 unnumbered interface-type interface-path-id
6. ipv6 enable
8. class-map type control subscriber match-any class-map_name
9. match protocol ppp
10. end-class-map
12. class-map type control subscriber match-all class-map_name
13. match protocol dhcpv6
14. end-class-map
18. class type control subscriber name do-all
20. end-policy-map
23. class type control subscriber name do-all
25. end-policy-map
28. description LINE
29. ipv6 enable
30. service-policy type control subscriber name
31. encapsulation dot1q 801
33. initiator dhcp
4.3.x
OL-28375-03 153
DHCPv6 Features
DETAILED STEPS
Example:
Step 1
Configures the dynamic template of type ppp and enters
the dynamic template type configuration mode.
dynamic-template type ppp dynamic_template_name
Example:
ppp ppp_pta_template
Step 2
Configures challenge handshake authentication protocol
(chap) and sets PPP link authentication method.
Example:
Step 3
Sets ipcp negotiation options and sets the peer address
configuration option for the peer-address pool.
ppp ipcp peer-address pool pool_name
Example:
ppp ipcp peer-address pool p1
Step 4
Enables IPv4 processing without an explicit address
for an interface.
ipv4 unnumbered interface-type interface-path-id
Example:
ipv4 unnumbered Loopback 1
Step 5
Example:
ipv6 enable
Step 6
Release 4.3.x
154 OL-28375-03
DHCPv6 Features
match-any criteria.
class-map type control subscriber match-any
class-map_name
Example:
subscriber match-any pta_class
Step 8
the earlier step.
match protocol ppp
Example:
ppp
Step 9
Example:
Step 10
match-all criteria.
class-map type control subscriber match-all class-map_name
Example:
subscriber match-all ipoe_test
Step 12
the earlier step.
Example:
dhcpv6
Step 13
Example:
Step 14
4.3.x
OL-28375-03 155
DHCPv6 Features
Configures the subscriber control policy map. policy-map type control subscriber policy_name
Example:
control subscriber policy1
Step 16
criteria.
Example:
Step 17
criteria.
class type control subscriber name do-all
Example:
RP/0/RSP0/CPU0:router(config-pmap)# class type
control subscriber ipoe_test1 do-all
Step 18
Step 19
Example:
dynamic-template v6_test1
Example:
Step 20
Configures the subscriber control policy map. policy-map type control subscriber policy_name
Example:
control subscriber policy1
Step 21
Configures the policy event with the match-all criteria. event session-start match-all
Example:
Step 22
criteria.
class type control subscriber name do-all
Example:
control subscriber pta_class do-all
Step 23
Release 4.3.x
156 OL-28375-03
DHCPv6 Features
Step 24
Example:
dynamic-template ppp_pta_template
Example:
Step 25
Configures an interface and enters the interface
configuration mode.
Example:
BundleEther1.1
Step 27
Sets the description for the above configured interface. description LINE
Example:
RP/0/RSP0/CPU0:router(config-if)# description IPoE
Step 28
Example:
RP/0/RSP0/CPU0:router(config-if)# ipv6 enable
Step 29
interface.
service-policy type control subscriber name
Example:
control subscriber ipoe1
Step 30
Enables encapsulated 802.1Q VLAN configuration. encapsulation dot1q 801
Example:
RP/0/RSP0/CPU0:router(config-if)# encapsulation dot1q
801
Step 31
4.3.x
OL-28375-03 157
DHCPv6 Features
Enables l2-connected IPv6 subscriber. ipsubscriber ipv6 l2-connected
Example:
l2-connected
Step 32
Configures IPv6 subscriber initiator. initiator dhcp
Example:
RP/0/RSP0/CPU0:router(config-if-ipsub-ipv6-l2conn)#
initiator dhcp
Step 33
Configuring IPv6 PPPoE Subscriber Interfaces: An example
configure
dynamic-template
type ppp PPP_PTA_TEMPLATE
ppp ipcp peer-address pool ADDRESS_POOL
ipv4 unnumbered Loopback0
ipv6 enable
!
type ipsubscriber v6_test1
ipv6 enable
!
!
class-map type control subscriber match-any PTA_CLASS
match protocol ppp
end-class-map
!
class-map type control subscriber match-all ipoe_test1
end-class-map
!
policy-map type control subscriber ipoe1
class type control subscriber ipoe_test1 do-all
24 activate dynamic-template v6_test1
!
!
end-policy-map
!
policy-map type control subscriber POLICY1
Release 4.3.x
158 OL-28375-03
DHCPv6 Features
class type control subscriber PTA_CLASS do-all
1 activate dynamic-template PPP_PTA_TEMPLATE
!
!
end-policy-map
!
description IPoE
ipv6 enable
service-policy type control subscriber ipoe1
initiator dhcp
Ambiguous VLAN Support
An Ambiguous VLAN is configured with a range or group of VLAN IDs. The subscriber sessions created
over ambiguous VLANs are identical to subscribers over regular VLANs that support all regular configurations
such as policy-map, VRFs, QoS, and ACL. Multiple subscribers can be created on a particular VLAN ID as
long as they contain a unique MAC address. Ambiguous VLANs enhance scalability by reducing the need
for configuring multiple access interfaces.
To enable DHCPv6 support, ambiguous VLANs are unnumbered on top of the bundle interface.
The ambiguous VLANs are named exactly the same way as regular VLANs. The ambiguous VLANs are
considered Layer 3 interfaces in contrast to EFP ranges allowed for l2transport interface.
Note
When DHCPv6 Server receives a SOLICIT message on the ambiguous VLAN interface, the VLAN IDs are
extracted fromthe received packet and used for authenticating the subscriber with the client related information.
When an interface configuration is changed from ambiguous to non-ambiguous or vice-versa or Ambiguous
VLAN range is changed, then all existing client bindings for the Ambiguous VLAN are cleared.
For more information on configuring ambiguous VLAN, see Configuring Ambiguous VLANs, on page 159.
Configuring Ambiguous VLANs
Perform this task to configure ambiguous vlans.
There is no DHCP-specific configuration required for ambiguous VLANs. Note
4.3.x
OL-28375-03 159
DHCPv6 Features
SUMMARY STEPS
1. configure
3. Use any of these encapsulations to configure encapsulated ambiguous VLANs:
encapsulation ambiguous { dot1q | dot1ad } {any | vlan-range }
encapsulation ambiguous dot1q vlan-id second-dot1q { any | vlan-range }
encapsulation ambiguous dot1q any second-dot1q { any | vlan-id }
encapsulation ambiguous dot1ad vlan-id dot1q { any | vlan-range }
4. ipv4 | ipv6address source-ip-address destination-ip-address
6. ipsubscriber { ipv4|ipv6 } l2-connected
7. initiator dhcp
DETAILED STEPS
Example:
Step 1
Configures the interface and enters the interface
configuration mode.
Example:
Bundle-Ether100.12
Step 2
Configures IEEE 802.1Q VLAN configuration. Use any of these encapsulations to configure encapsulated
ambiguous VLANs:
Step 3
The vlan-range can be given in comma-separated, or
hyphen-separated format, or a combination of both,
as shown in the examples.
encapsulation ambiguous { dot1q | dot1ad } {any |
vlan-range }
encapsulation ambiguous dot1q vlan-id second-dot1q {
any | vlan-range }
encapsulation ambiguous dot1q any second-dot1q { any
| vlan-id }
encapsulation ambiguous dot1ad vlan-id dot1q { any |
vlan-range }
Release 4.3.x
160 OL-28375-03
DHCPv6 Features
Example:
RP/0/RSP0/CPU0:router(config-if)# encapsulation
ambiguous dot1q any
ambiguous dot1q 14 second-dot1q 100-200
ambiguous dot1q any second-dot1q any
ambiguous dot1ad 14 dot1q 100,200,300-400
Configures the IPv4 or IPv6 protocol address. ipv4 | ipv6address source-ip-address destination-ip-address
Example:
RP/0/RSP0/CPU0:router(config-if)# ipv4 address 2.1.12.1
Step 4
255.255.255.0
RP/0/RSP0/CPU0:router(config-if)# ipv6 address 1:2:3::4
128
Applies a policy-map to an access interface where the
policy-map was previously defined with the specified
PL1 policy_name.
Example:
control subscriber PL1
Step 5
Enables l2-connected IPv4 or IPv6 IP subscriber. ipsubscriber { ipv4|ipv6 } l2-connected
Example:
l2-connected
l2-connected
Step 6
Enables initiator DHCP on the IP subscriber. initiator dhcp
Example:
RP/0/RSP0/CPU0:router(config-if)# initiator dhcp
Step 7
YesSaves configuration changes and exits the
4.3.x
OL-28375-03 161
DHCPv6 Features
Configuring Ambiguous VLANs: An example
configure
encapsulation ambiguous dot1q 14 second-dot1q any
ipv4 address 2.1.12.1 255.255.255.0
initiator dhcp
!
!
end
DHCPv6 Address or Prefix Pool
An address or prefix pool represents a pool of available address or prefixes from which a delegating router
assigns an address or delegates a prefix to the requesting router. The Distributed Address Pool Service (DAPS)
manages and maintains address or prefix pools for DHCPv6.
DHCPv6 Prefix Delegation involves a delegating router selecting a prefix and delegating it on a temporary
basis to a requesting router. The delegating router assigns the address or delegates the prefix from the address
pool or prefix pool to the requesting router.
For more information about configuring DHCPv6 address or prefix pool, see Configuring IPv6 Address or
Prefix Pool Name, on page 162.
Configuring IPv6 Address or Prefix Pool Name
Perform this task to configure IPv6 address or prefix pool name under dynamic template configuration mode.
SUMMARY STEPS
1. configure
2. dynamic-template
3. type ipsubscriber dynamic-template_name
4. dhcpv6 delegated-prefix-pool pool-name
6. type ppp dynamic-template_name
7. dhcpv6 address-pool pool-name
10. dhcpv6 address-pool pool-name
12. ipv6 nd framed-prefix-pool pool-name
Release 4.3.x
162 OL-28375-03
DHCPv6 Features
DETAILED STEPS
Example:
Step 1
Enables dynamic template configuration. dynamic-template
Example:
Step 2
Configures dynamic template of type ipsubscriber and
enters the dynamic-template type configuration mode.
type ipsubscriber dynamic-template_name
Example:
ipsubscriber ipv6-sub-template
Step 3
Configures IPv6 subscriber dynamic template with
prefix-delegation pool.
dhcpv6 delegated-prefix-pool pool-name
Example:
dhcpv6 delegated-prefix-pool mypool
Step 4
Configures dynamic template of type ppp. type ppp dynamic-template_name
Example:
ppp ipv6-sub-template
Step 6
Configures IPv6 address pool for PPPoE subscribers. dhcpv6 address-pool pool-name
Example:
dhcpv6 address-pool my-pppoe-addr-pool
Step 7
4.3.x
OL-28375-03 163
DHCPv6 Features
enters the dynamic-template type configuration mode.
Example:
ipsubscriber my-ipv6-template
Step 9
Configures IPv6 address pool for IPoE subscribers. dhcpv6 address-pool pool-name
Example:
dhcpv6 address-pool my-ipsub-addr-pool
Step 10
Configures prefix pool to be used by SLAAC only. ipv6 nd framed-prefix-pool pool-name
Example:
framed-prefix-pool my-slaac-pool
Step 12
Release 4.3.x
164 OL-28375-03
DHCPv6 Features
Configuring IPv6 Address or Prefix Pool Name: An example
configure
dynamic-template
type ipsubscriber ipv6-sub-template
dhcpv6 delegated-prefix-pool mypool
end
dynamic-template
type ppp ipv6-sub-template
dhcpv6 address-pool my-pppoe-addr-pool
!
type ipsubscriber my-ipv6-template
dhcpv6 address-pool my-ipsub-addr-pool
!!
ipv6 nd framed-prefix-pool my-slaac-pool
end
!!
DHCPv6 Dual-Stack Lite Support
Dual-Stack Lite (DS-Lite) is a technique for providing complete support for both IPv4 and IPv6 internet
protocols, both in hosts and router. Dual-Stack Lite enables a broadband service provider to share IPv4
addresses among customers by combining two technologies: IP in IP (IPv4- in-IPv6) and Network Address
Translation (NAT).
The DS-Lite feature contains two components: Basic Bridging Broad Band ( B4) and Address Family Transition
Router (AFTR).
The B4 element is a function implemented on a dual-stack-capable node, either a directly connected device
or a CPE that creates a tunnel to an Address Family Transition Router (AFTR). On the other hand, an AFTR
element is the combination of an IPv4-in-IPv6 tunnel endpoint and an IPv4-IPv4 NAT implemented on the
same node. A DS-Lite B4 element uses a DHCPv6 option to discover the IPv6 address of its corresponding
AFTR location.
For more information about configuring AFTR for DS-Lite, see Configuring AFTR Fully Qualified Domain
Name for DS-Lite, on page 165.
Configuring AFTR Fully Qualified Domain Name for DS-Lite
Perform this task to configure AFTR fully qualified domain name for DS-Lite.
SUMMARY STEPS
1. configure
2. dhcp ipv6
4. aftr-name aftr_name
4.3.x
OL-28375-03 165
DHCPv6 Features
DETAILED STEPS
Example:
Step 1
configuration mode.
dhcp ipv6
Example:
Step 2
Configures DHCPv6 server profile and enters the
Example:
Step 3
Configures the AFTR Fully Qualified Domain Name
option, in the server profile mode, for the DS-Lite
support.
aftr-name aftr_name
Example:
aftr-name aftr-server.example.com
Step 4
Configuring AFTR Fully Qualified Domain Name for DS-Lite: An example
configure
dhcp ipv6
aftr-name aftr-server.example.com
end
!!
Release 4.3.x
166 OL-28375-03
DHCPv6 Features
VRF Awareness in DHCPv6
VRF Awareness is the ability of DHCPv6 Server or Proxy to support multiple clients in different VPNs where
the same IP address is assigned to clients on differing VPNs. The IPv6 addresses in a VRF is independent
from IPv6 addresses in an another VRF. It is not mandatory to have same prefix/address in multiple VRFs.
For more information about defining VRF in a dynamic template, see Defining VRF in a Dynamic Template,
on page 167.
Defining VRF in a Dynamic Template
Perform this task for defining VRF in a dynamic template. The IPv6 addresses in a VRF is independent from
IPv6 addresses in an another VRF. It is not mandatory to have same prefix or address in multiple VRFs.
SUMMARY STEPS
1. configure
2. dynamic-template
4. vrf vrf_name
DETAILED STEPS
Example:
Step 1
Enables dynamic template configuration. dynamic-template
Example:
Step 2
enters the dynamic template type configuration mode.
Example:
ipsubscriber ipv6-sub-template
Step 3
Sets the VRF in which the interface operates. vrf vrf_name
Example:
vrf vrf1
Step 4
4.3.x
OL-28375-03 167
DHCPv6 Features
Defining VRF in a Dynamic Template: An example
configure
dynamic-template
type ipsubscriber ipv6-sub-template
vrf vrf1
end
!!
Packet Handling on Subscriber Interfaces
This section describes how subscriber interfaces are supported in certain special cases. These special cases
include L3 forwarded interfaces. As a result, this support is applicable only to PPP over Ethernet PPP
Termination and Aggregation (PPPoE PTA) and IPoE sessions.
Most subscriber data packets are forwarded directly by the network processing unit (NPU). There are certain
special cases where the NPU does not completely handle the data packet. These special cases are handled by
the CPU, and go through an internal interface created for this purpose. This internal interface is named the
Subscriber Interface or SINT. SINT is an aggregate interface, which is used by all packets punted on subscriber
interfaces. There is one SINT for each node. When the BNG package is installed, by default the SINT is
created. The SINT interfaces are needed for punt-inject of packets on subscriber interfaces.
These special cases are supported for both IPoE and PPPoE PTA:
These special cases do not apply to PPPoE L2TP, because it is an L2 service. Note
Ping to and from subscriber
BNG allows the receiving of a ping request from both IPoE and PPPoE PTA subscriber interfaces; this
is consistent with other non-BNG interface types as well. Similarly, BNG also allows the sending of a
ping request to both IPoE and PPPoE PTA subscriber interfaces. This includes:
various lengths of ping packets including lengths exceeding the subscribers MTU size
subscriber in the default and private VRFs
various ping options such as type of service, DF set, and verbose
BNG also supports receiving a ping request from both IPv4 and IPv6 subscribers.
Release 4.3.x
168 OL-28375-03
Excessive Punt Flow Trap feature should be disabled when sending a high rate of pings
to, or from subscriber interfaces.
Note
Option Handling
BNG supports handling IP options; this is consistent with non-BNG interface types. These are punted
from the NPU to the CPU. These go through the SINT interface and are handled by the appropriate
application.
Support for traceroute, PMTU discovery, ICMP unreachable
BNG supports sending ICMP for packets that are received from or destined to a PPPoE or IP
subscriber interface that cannot be forwarded. This functionality is similar to other non-BNG
subscriber interfaces.
BNG supports PMTU, in which BNG sends ICMPs, when a packet is destined to a subscriber
interface, but the packet exceeds the subscriber MTU and the DF bit is set.
BNGsupports sending ICMPs when packets to (egress ACL) or from(ingress ACL) the subscriber
interface are denied due to the ACL. If the ACL is configured do both deny and log, then the
packets get dropped, but no ICMP is generated.
BNG supports traceroute functionality that enables sending an ICMP when the time to live (TTL)
of the packet is exceeded.
BNG supports traceroute functionality for both IPv4 and IPv6 subscribers.
Fragmentation
BNG does not support fragmentation of packets destined to the PPPoE or IP subscriber interfaces.
In Cisco IOS XR, fragmentation is handled by linecard (LC) CPU or route processor
(RP) CPU. All packets requiring fragmentation are policed by local packet transport
service (LPTS), to a maximum of 2500 packets per second (pps) for each network
processing unit (NPU).
The fragmentation path is supported only in software, and fragmented packets skip all
features, including subscriber features, QoS, ACL and so on. Therefore, irrespective of
BNG, it should not be used as a general forwarding path.
Caution
Restrictions
These restrictions apply to implementing subscriber interfaces:
During an ACL logging, packets are punted to CPU, and BNG interfaces are directed to the SINT
interface. The SINT interface drops these log packets because the systemdoes not support ACL Logging
on BNG interfaces.
IPv6 Ping and traceroute functions should use both the CPE and BNG routers global addresses. IPv6
Ping and traceroute functions using link local address does not work in all cases.
Logging on subscriber ACLs is not supported.
4.3.x
OL-28375-03 169
IPv6 Neighbor Discovery
The IPv6 neighbor discovery process uses Internet Control Message Protocol (ICMP) messages and
solicited-node multicast addresses to determine the link-layer address of a neighbor on the same network
(local link), verify the reachability of a neighbor, and track neighboring routers.
The IPv6 static cache entry for neighbor discovery feature allows static entries to be made in the IPv6 neighbor
cache. Static routing requires an administrator to manually enter IPv6 addresses, subnet masks, gateways, and
corresponding Media Access Control (MAC) addresses for each interface of each device into a table. Static
routing enables more control but requires more work to maintain the table. The table must be updated each
time routes are added or changed.
The different message types in neighbor discovery are:
IPv6 Neighbor Solicitation Message: A value of 135 in the Type field of the ICMP packet header
identifies a neighbor solicitation message. Neighbor solicitation messages are sent on the local link when
a node wants to determine the link-layer address of another node on the same local link.
IPv6 Router Advertisement Message: Router advertisement (RA) messages, which have a value of 134
in the Type field of the ICMP packet header, are periodically sent out each configured interface of an
IPv6 device.
Ambiguous VLAN does not have association with any particular VLAN, and therefore, a unicast router
advertisement message has to be sent out for ambiguous VLANinterfaces. To enable IPv6 unicast router
advertisement, you must use the ipv6 nd ra-unicast command in the dynamic template configuration
mode.
IPv6 Neighbor Redirect Message: A value of 137 in the type field of the ICMP packet header identifies
an IPv6 neighbor redirect message. Devices send neighbor redirect messages to inform hosts of better
first-hop nodes on the path to a destination.
In BNG, IPv6 neighbor discovery supports both IPoE and PPPoE sessions. IPv6 neighbor discovery provides
Stateless Address Auto Configuration (SLAAC), which is used for assigning a prefix to the PPPoE subscriber.
These sections provide references related to implementing PPP, PPPoE, L2TP, and DHCP.
RFCs
Title Standard/RFC - PPP
The PPP Internet Protocol Control Protocol (IPCP) RFC-1332
PPP LCP Extensions RFC-1570
The Point-to-Point Protocol (PPP) RFC-1661
PPP Challenge Handshake Authentication Protocol
(CHAP)
RFC-1994
Release 4.3.x
170 OL-28375-03
IPv6 Neighbor Discovery
Title Standard/RFC - PPPoE
A Method for Transmitting PPP Over Ethernet
(PPPoE)
RFC-2516
DSL Forum Vendor-Specific RADIUS Attributes RFC-4679
Title Standard/RFC - L2TP
Layer two tunneling protocol "L2TP" RFC-2661
MIBs
MIBs Link MIBs
Link Description
http://www.cisco.com/cisco/web/support/index.html The Cisco Technical Support website contains
thousands of pages of searchable technical content,
including links to products, technologies, solutions,
technical tips, and tools. Registered Cisco.com users
can log in fromthis page to access even more content.
4.3.x
OL-28375-03 171
Release 4.3.x
172 OL-28375-03
C HAP T E R 6
Deploying the Quality of Service (QoS)
The Quality of Service (QoS) feature ensures that traffic to and from priority applications gets preference
in using network resources. QoS actions are defined by service-policies that are deployed using policy-maps.
During the QoS process, packets are encapsulated with QoS information. The encapsulation is monitored
and accounted by the QoS accounting function.
Parameterized QoS (PQoS) is another form of QoS in which the traffic priority is based on the characteristic
of the data being carried by the traffic.
BNG supports merging of multiple QoS policy-maps applied through multiple dynamic templates and
implementing them on a single subscriber.
This chapter explains deploying QoS, and covers the following topics:
Quality of Service Overview, page 173
Parameterized QoS, page 178
QoS Accounting, page 189
Support for Shared Policy Instance, page 191
Merging QoS Policy-maps, page 198
QoS Features Supported on BNG, page 202
Quality of Service Overview
Quality of Service (QoS) is the technique of prioritizing network traffic for time-sensitive and mission-critical
applications such as VoIP services, live streaming of videos, priority accesses to database, and so on. Functions
that QoS provides ensure that such applications receive sufficient bandwidth at low latency, with reduced
data loss.
QoS functions perform preferential forwarding of high-priority packets. To do so, the packet is identified,
classified, and then prioritized on all routers along the data-forwarding path throughout the network. As a
result, priority applications get the resources they require, while other applications access the network,
simultaneously.
4.3.x
OL-28375-03 173
QoS functions provide service providers cost benefits by enabling them to use existing resources efficiently
and ensure the required level of service without reactively expanding, or over-provisioning their networks.
QoS also improves customer experience when they get reliable services froma variety of network applications.
It is ideal to deploy QoS on BNG because BNG is present at the edge router, and subscriber directly connects
to it. One of the unique features of BNG is QoS accounting. This feature enables BNG to collect and report
QoS encapsulation information to the RADIUS server. For details, see QoS Accounting, on page 189.
The deployment of QoS involves three components:
Class-map Classifies different forms of traffic, like video, data, VOIP, etc, based on matching rules.
Policy-map Defines the QoS actions to be applied to the classified traffic. It references the classes
previously defined in the class-map. These policy-maps are also called QoS maps. The actions defined
in the policy-map perform traffic prioritization and bandwidth allocation.
Service policy Associates a previously defined policy-map with a attachment point and direction, on
BNG. The attachment points are listed in the section QoS Attachment Points, on page 205.The two
directions possible for a policy is input and output. The policy direction is relative to the attachment
point.
BNG supports two-level hierarchal policy (parent policy and child policy) for deploying QoS. Based on the
preference of service provider, the QoS policies are defined and applied on BNG in these ways:
Define and apply the QoS policy from CLI. See, Configuring Service-policy and Applying Subscriber
Settings Through Dynamic Template, on page 176.
Define the QoS policy in CLI, but apply it fromRAIDUS. See, Configuring Service-policy and Applying
Subscriber Settings Through RADIUS, on page 174.
Define and apply the QoS policy from RAIDUS. It is also called Parameterized QoS, on page 178.
Restriction
If the subscriber ingress or egress QOS includes policing, shaping, bandwidth, or WRED actions, it is
recommended that only active:standby bundle interfaces be used. Load-sharing should be avoided.
Configuring Service-policy and Applying Subscriber Settings Through RADIUS
Perform this task to deploy the QoS policy using CLI commands. In this task, subscriber settings are applied
from the RADIUS server.
SUMMARY STEPS
1. configure
2. policy-map type qos q_in
3. class class-default
4. service-policy q_child_in
5. policy-map type qos q_out
7. service-policy q_child_out
Release 4.3.x
174 OL-28375-03
DETAILED STEPS
Example:
Step 1
Configures the policy-map for the type qos. policy-map type qos q_in
Example:
RP/0/RSP0/CPU0:router(config)# policy-map
type qos q_in
Step 2
Configures or modifies the parent class-default class. class class-default Step 3
Example:
RP/0/RSP0/CPU0:router(config-pmap)# class
class-default
You can configure only the class-default class in a parent
policy. Do not configure any other traffic class.
Note
Applies a bottom-level policy to the top-level class-default class. service-policy q_child_in
Example:
RP/0/RSP0/CPU0:router(config-pmap-c)#
service-policy q_child_in
Step 4
Configures the policy-map for the type qos. policy-map type qos q_out
Example:
type qos q_out
Step 5
Example:
class-default
You can configure only the class-default class in a parent
policy. Do not configure any other traffic class.
Note
Applies a bottom-level policy to the top-level class-default class. service-policy q_child_out
Example:
service-policy q_child_out
Step 7
4.3.x
OL-28375-03 175
Configuring Subscriber Policy through CLI and Applying through RADIUS: Examples
configure
policy-map type qos q_in
class class-default
end
\\the following procedure is ran in RADIUS
Service-Type = Outbound-User
Cisco-avpair = "ip:keepalive=protocol arp attempts 5 interval 15",
Cisco-avpair = "ipv4:ipv4-mtu=750",
Cisco-avpair = "ipv4:ipv4-unnumbered=Loopback0",
Cisco-avpair = "subscriber:sub-qos-policy-in=q_in",
Cisco-avpair = "subscriber:sub-qos-policy-out=q_out",
Idle-Timeout = 1000,
Session-Timeout = 5000
Configuring Service-policy and Applying Subscriber Settings Through Dynamic
Template
Perform this task to deploy the QoS policy using CLI commands. In this task, subscriber settings are applied
using a dynamic template.
SUMMARY STEPS
1. configure
2. policy-map type qos q_in
4. service-policy q_child_in
5. policy-map type qos q_out
7. service-policy q_child_out
8. dynamic-template type ppp dynamic_config
9. service-policy input q_in
10. service-policy output q_out
Release 4.3.x
176 OL-28375-03
Configuring Service-policy and Applying Subscriber Settings Through Dynamic Template
DETAILED STEPS
Example:
Step 1
Configures the policy-map in the input direction. policy-map type qos q_in
Example:
RP/0/RSP0/CPU0:router(config)# policy-map type qos
q_in
Step 2
Example:
class-default
You can configure only the class-default class
in a parent policy. Do not configure any other
traffic class.
Note
Configures the service policy for the input direction. service-policy q_child_in Step 4
Example:
RP/0/RSP0/CPU0:router(config)# service-policy
q_child_in
The q_in and q_out policy maps are parent
policy maps.
Note
Configures the policy-map for the output direction. policy-map type qos q_out Step 5
Example:
RP/0/RSP0/CPU0:router(config)# policy-map type qos
q_out
policy maps.
Note
Example:
class-default
You can configure only the class-default class
in a parent policy. Do not configure any other
traffic class.
Note
Applies a bottom-level policy to the top-level
class-default class.
service-policy q_child_out
Example:
q_child_out
Step 7
policy maps.
Note
Configures dynamic-template of the type ppp and applies
the configuration through dynamic-template.
dynamic-template type ppp dynamic_config
Example:
ppp dynamic_config
Step 8
4.3.x
OL-28375-03 177
Configuring Service-policy and Applying Subscriber Settings Through Dynamic Template
Configures the service-policy in the input direction. service-policy input q_in
Example:
service-policy input q_in
Step 9
Configures the service-policy in the output direction. service-policy output q_out
Example:
service-policy input q_out
Step 10
Configuring Subscriber Policy through CLI and Applying to Subscriber through Dynamic-Template: Examples
configure
policy-map type qos q_in // policy-map input direction
class class-default
end
configure
policy-map type qos q_out // policy-map output direction
class class-default
end
// applying configuration through dynamic-template
configure
dynamic-template type ppp dynamic_policy
service-policy input q_in
service-policy output q_out
end
Parameterized QoS
Parameterized Quality of Service (PQoS) guarantees reliable performance of a network application by reserving
for it the required network bandwidth. In this case, the prioritization is based on the type of data being carried
by the packet.
Release 4.3.x
178 OL-28375-03
Parameterized QoS
In the standard QoS, the importance of a packet is based on the priority level that is defined for it. It is possible
that in once case a video packet and an asynchronous data transfer packet have the same priority level defined.
In such a case, the router gives equal importance to both packets. As a result, because of bandwidth conflict,
there can be video degradation.
On the other hand, in PQoS, packet importance is based on the characteristics or parameters of the data that
is carried by the packet. For example, it is possible to have PQoS provide dedicated bandwidth for video
packets. Even at times when heavy loads of asynchronous data traffic are introduced into the network, PQoS
guarantees that video packets have priority over other data streams that do not require real-time streaming.
Parameterized QoS has the ability to define, modify, or delete QoS policy-map based Vendor Specific Attributes
(VSAs). VSAs are downloaded through the RADIUS server. The attributes from the parameterized QoS
policies are filtered and passed on to the policy object library; the latter parses and translates them into policy
objects. The VSAs define a two-level hierarchical policy to be applied on the subscriber session. The format
of the QoS VSAs is:
AVPair: qos-policy-in=add-class(sub,<parent-class, child-class>,<action-list>)
AVPair: qos-policy-out=add-class(sub,<parent-class, child-class>,<action-list>)
AVPair: qos-policy-in=remove-class(sub,<parent-class, child-class>)
AVPair: qos-policy-out=remove-class(sub,<parent-class, child-class>)
where:
sub, is a constant string, signifies that the current policy on the subscriber is to be modified
<class-list> gives the hierarchy of the class to be added or removed (i.e. parent-class, child-class)
<action-list> gives the QoS actions to be applied under the class being added
For more information about QoS parameters and its syntax, see Parameterized QoS Syntax in the Configuring
Parameterized QoS Policy Through RADIUS, on page 184.
When a parameterized QoS policy for a subscriber is downloaded from the RADIUS server for the first time,
the VSAs are used to build the policy from scratch. After the policy is applied on the subscriber, any new or
modified VSAs downloaded for that subscriber from the RADIUS server automatically modifies the already
applied policy.
For deploying a Parameterized QoS policy from the RADIUS server, see Configuring Parameterized QoS
Policy Through RADIUS, on page 184.
Using Change of Authorization (CoA), it is possible to update the service-policy by modifying the class-maps
that were previously configured by the parameterized QoS. Modifying can involve removing existing classes,
or adding new classes. To make updates to the service-policy, see Modifying Service Policy through CoA ,
on page 187.
Parameterized QoS Syntax
Commands Qualifiers QoS Action
Parameter
shape(<rate-in-kbps>) QoS Action Shape
shape average <shape-rate> <kbps> CLI Equivalent
4.3.x
OL-28375-03 179
Parameter
qos-policy-out:add-class(sub,(class-default),shape(14700)) RADIUS
Equivalent -
Example
Shape-rpct(<rate-in-pct>) QoS Action Shape in
percentage
shape average percent < rate-in-pct > CLI Equivalent
qos-policy-out:add-class(sub,(class-default),shape-pct(25)) RADIUS
Equivalent -
Example
police( <conform-rate-in-kbps>,
<conform-burst-in-kBytes>,
<exceed-rate-in-kbps>,
<exceed-burst-in-kbytes>,
<conform-action>,
<exceed-action>,
<violate-action>)
QoS Action Police (Variant
1)
police rate <conform-rate> <kbps> burst <conform-burst> <kbps>
peak-rate <exceed-rate>
exceed-burst <exceed-burst>
conform-action <action>
exceed-action <action>
violate-action <action>
CLI Equivalent
qos-policy-in:add-class(sub,(class-default, voip),police(2000,2000,
4000, 4000,transmit, set-ipprec(<
precedence>), drop) )
RADIUS
Equivalent -
Example
Police (<conform-rate-in-kbps>) QoS Action Police (Variant
2)
police rate <kbps> CLI Equivalent
qos-policy-in:add-class(sub,(class-default, voip), police(200000) ) RADIUS
Equivalent -
Example
Release 4.3.x
180 OL-28375-03
Parameter
police-rpct(<conform-rate-in-pct>,
<conform-burst-in-us>,
<exceed-rate-in-pct>,
<exceed-burst-in-us>,
<conform-action>,
<exceed-action>,
<violate-action>)
QoS Action Police in
percentage
(Variant 1)
police rate percentage <pct> burst <conform-burst> < us> peak-rate
percentage<pct> exceedburst
<exceed-burst>
conform-action <action>
exceed-action <action>
violate-action <action>
CLI Equivalent
qos-policy-in:add-class(sub,(class-default, voip),police-rpct(20,20, 40,
40,transmit, set-ipprec(<
precedence>), drop) )
RADIUS
Equivalent -
Example
Police-rpct(<conform-rate-in-pct> QoS Action Police in
percentage
(Variant 2)
police rate percentage <pct> CLI Equivalent
qos-policy-in:add-class(sub,(class-default, voip), police-rpct(20) ) RADIUS
Equivalent -
Example
set-ip-prec(<precedence>) QoS Action Set IP
Precedence
set precedence <precedence> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip), set-ip-prec(5)) RADIUS
Equivalent -
Example
set-cos(<cos-val>) QoS Action Set CoS
set cos <cos-val> CLI Equivalent
4.3.x
OL-28375-03 181
Parameter
qos-policy-out:add-class(sub,(class-default,voip), set-cos(5)) RADIUS
Equivalent -
Example
bw-abs(<bw-in-kbps>) QoS Action Minimum
Bandwidth
bandwidth <bw-in-kbps> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,video),bw-abs(2000)) RADIUS
Equivalent -
Example
bw-pct(<bw-in-pct>) QoS Action Minimum
bandwidth
percentage
bandwidth percent <pct> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,video),bw-abs(2000)) RADIUS
Equivalent -
Example
bw-rpct(<pct>) QoS Action Bandwidth
Remaining
Percentage
bandwidth remaining percent <pct> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip),bw-rpct(33)) RADIUS
Equivalent -
Example
set-ip-dscp(<dscp-val>) QoS Action Set IP DSCP
Set dscp <dscp-val> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip), set-ip-dscp(46)) RADIUS
Equivalent -
Example
queue-limit(<qlimit-in-packets>) QoS Action Queue Limit in
packets
queue-limit <val> < packets> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip),queue-limit(64)) RADIUS
Equivalent -
Example
Release 4.3.x
182 OL-28375-03
Parameter
queue-limit-us(<qlimit-in-us>) QoS Action Queue Limit in
us
queue-limit <val> <us> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip),queue-limit-us(240)) RADIUS
Equivalent -
Example
random-detect-dscp(<dscp>, <min-thresold>, <max-threshold>,
<probability>)
QoS Action DSCP based
WRED
random-detect dscp <dscp-val> < Min-thresh> <Kbytes> <max-thresh>
<Kbytes> probability <
probability-val>
CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip), random-detect-dscp
(24, 25000, 35000))
RADIUS
Equivalent -
Example
random-detect-prec (<precedence>, <min-thresold>, <max-threshold>,
<probability>)
QoS Action Precedence
based WRED
random-detect precedence <prec-val> < Min-thresh> <Kbytes>
<max-thresh> <Kbytes>
probability < probability-val>
CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip), random-detect- (24,
25000, 35000))
RADIUS
Equivalent -
Example
set-qos-grp(<group-val>) QoS Action Set qos group
set qos-group <qos-group-val> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip), set-qos-grp (24)) RADIUS
Equivalent -
Example
pri-level(<priority-level>) QoS Action Priority Level
priority level <priority-level> CLI Equivalent
qos-policy-out:add-class(sub,(class-default, voip), pri_level(1)) RADIUS
Equivalent -
Example
set-dclass(<discard-class-val>) QoS Action Set discard class
4.3.x
OL-28375-03 183
Parameter
set discard-class <discard-class-val> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip), set-dclass (4)) RADIUS
Equivalent -
Example
set-mpls-exp-topmost (<mpls-exp- topmost-val>) QoS Action Set MPLS exp
topmost bit
set mpls experimental topmost <mpls-exp- topmost-val> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip), set-mpls-exp-topmost
(4))
RADIUS
Equivalent -
Example
set-mpls-exp- imposition (<mpls-exp-imposition-val>) QoS Action Set MPLS exp
imposition bit
set mpls experimental imposition <mpls-exp- imposition-val> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip), set-mpls-exp-
imposition (4))
RADIUS
Equivalent -
Example
set-tunnel-prec(<prec-val>) QoS Action Set Tunnel
precedence
set precedence tunnel <precedence-val> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip), set-tunnel-prec(4)) RADIUS
Equivalent -
Example
set-tunnel-dscp (<dscp-val>) QoS Action Set Tunnel
DSCP
set dscp tunnel <dscp-val> CLI Equivalent
qos-policy-out:add-class(sub,(class-default,voip), set-tunnel-dscp(4)) RADIUS
Equivalent -
Example
Configuring Parameterized QoS Policy Through RADIUS
Perform this task to deploy parameterized QoS policy and apply subscriber settings through the RADIUS
server. These steps are performed on the RADIUS server for each subscriber or service profile.
Release 4.3.x
184 OL-28375-03
In parameterized QoS configuration, the policy-map is not defined on the CLI. It is dynamically created
based on the configuration passed through RADIUS. This procedure applies to the RADIUS server as
part of RADIUS user configurations. The policy-map results are applied to the subscriber when that user
profile is downloaded after executing a control policy authentication or authorization action. The class-map
must be configured through CLI. For this task, the classes voice_in, video_in, data_in, video_out, voice_out,
and data_out are configured separately.
Note
SUMMARY STEPS
1. Cisco-AVPair = "ip:qos-policy-in=add-class(sub, (class-default),police(2000))"
2. Cisco-AVPair+= "ip:qos-policy-in=add-class(sub, (class-default,voice_in), pri-level(1), police(256))"
3. Cisco-AVPair+= ip:qos-policy-in=add-class(sub, (class-default,video_in), pri-level(2), police(1000))"
4. Cisco-AVPair += "ip:qos-policy-in=add-class(sub, (class-default,data_in), set-qos-grp(4))"
5. Cisco-AVPair += "ip:qos-policy-in=add-class(sub, (class-default,class-default), set-qos-grp(7))"
6. Cisco-AVPair+= "ip:qos-policy-out=add-class(sub, (class-default), shape(4000))"
7. Cisco-AVPair += "ip:qos-policy-out=add-class(sub, (class-default,voice_out),
pri-level(1),queue-limit-us(10000))"
8. Cisco-AVPair+= "ip:qos-policy-out=add-class(sub, (class-default,video_out),queue-limit-us(30000),
shape(2000))"
9. Cisco-AVPair+= "ip:qos-policy-out=add-class(sub, (class-default,data_out), bw-rpct(20))"
10. Cisco-AVPair+= "ip:qos-policy-out=add-class(sub, (class-default,class-default))"
DETAILED STEPS
Configures the cisco-avpair class-map in input
direction for police action parameter.
Cisco-AVPair = "ip:qos-policy-in=add-class(sub,
(class-default),police(2000))"
Example:
(class-default),police(2000))"
Step 1
direction for the police action parameter.
Cisco-AVPair+= "ip:qos-policy-in=add-class(sub,
(class-default,voice_in), pri-level(1), police(256))"
Example:
(class-default,voice_in), pri-level(1), police(256))"
Step 2
Cisco-AVPair+= ip:qos-policy-in=add-class(sub,
(class-default,video_in), pri-level(2), police(1000))"
Example:
Cisco-AVPair = ip:qos-policy-in=add-class(sub,
(class-default,video_in), pri-level(2), police(1000))"
Step 3
4.3.x
OL-28375-03 185
Cisco-AVPair += "ip:qos-policy-in=add-class(sub,
(class-default,data_in), set-qos-grp(4))"
Example:
(class-default,data_in), set-qos-grp(4))"
Step 4
direction for the set qos action parameter.
Cisco-AVPair += "ip:qos-policy-in=add-class(sub,
(class-default,class-default), set-qos-grp(7))"
Example:
(class-default,class-default), set-qos-grp(7))"
Step 5
Configures the cisco-avpair class-map in output
direction for the shape action parameter.
Cisco-AVPair+= "ip:qos-policy-out=add-class(sub, (class-default),
shape(4000))"
Example:
Cisco-AVPair = "ip:qos-policy-out=add-class(sub,
(class-default), shape(4000))"
Step 6
direction for the queue-limit-us action parameter.
Cisco-AVPair += "ip:qos-policy-out=add-class(sub,
(class-default,voice_out), pri-level(1),queue-limit-us(10000))"
Example:
Step 7
(class-default,voice_out),
direction for the queue-limit-us and the shape action
parameters.
Cisco-AVPair+= "ip:qos-policy-out=add-class(sub,
(class-default,video_out),queue-limit-us(30000), shape(2000))"
Example:
Step 8
(class-default,video_out),queue-limit-us(30000),
shape(2000))"
direction for the bandwidth action parameter.
(class-default,data_out), bw-rpct(20))"
Example:
(class-default,data_out), bw-rpct(20))"
Step 9
direction for the class action parameter.
(class-default,class-default))"
Step 10
Example:
(class-default,class-default))"
Release 4.3.x
186 OL-28375-03
For the complete list of QoS action
parameters that can be configured and
applied through RADIUS, see
Parameterized QoS Syntax section in
Parameterized QoS Syntax, on page 179.
Note
Configuring Parameterized Subscriber Policy Defined and Applied through RADIUS: An example
Cisco-AVPair = "ip:qos-policy-in=add-class(sub, (class-default),police(2000))"
Cisco-AVPair += "ip:qos-policy-in=add-class(sub, (class-default,voice_in), pri-level(1),
police(256))"
Cisco-AVPair += "ip:qos-policy-in=add-class(sub, (class-default,video_in), pri-level(2),
police(1000))"
Cisco-AVPair += "ip:qos-policy-in=add-class(sub, (class-default,data_in), set-qos-grp(4))"
Cisco-AVPair += "ip:qos-policy-in=add-class(sub, (class-default,class-default),
set-qos-grp(7))"
Cisco-AVPair += "ip:qos-policy-out=add-class(sub, (class-default), shape(4000))"
Cisco-AVPair += "ip:qos-policy-out=add-class(sub, (class-default,voice_out),
Cisco-AVPair += "ip:qos-policy-out=add-class(sub,
(class-default,video_out),queue-limit-us(30000), shape(2000))"
Cisco-AVPair += "ip:qos-policy-out=add-class(sub, (class-default,data_out), bw-rpct(20))"
Cisco-AVPair += "ip:qos-policy-out=add-class(sub, (class-default,class-default))"
Modifying Service Policy through CoA
Perform this task to modify service-policy through CoA.
The Web Portal or Radius server that supports CoA should be configured to generate a CoA request with
Cisco VSA corresponding to the steps in this task.
Note
SUMMARY STEPS
1. qos-policy-out remove-class(sub, (class-default, voip))
2. qos-policy-out add-class(sub, (class-default, video), bw-rpct(50), pri-level(2))
3. qos-policy-out add-class(sub, (class-default, data), shape(400),set-ip-prec(1))
4.3.x
OL-28375-03 187
DETAILED STEPS
Removes the class map, where voip is the class to be
removed from a previously configured parameterized
QoS for a subscriber.
qos-policy-out remove-class(sub, (class-default, voip))
Example:
qos-policy-out=remove-class(sub, (class-default,
voip))
Step 1
Adds a class map, where video is the class to be added
to a previously configured parameterized QoS for a
subscriber.
qos-policy-out add-class(sub, (class-default, video),
bw-rpct(50), pri-level(2))
Example:
qos-policy-out=add-class(sub, (class-default, video),
bw-rpct(50), pri-level(2))
Step 2
Configures the qos-policy-out for shape, set ip
precedence parameters.
qos-policy-out add-class(sub, (class-default, data),
shape(400),set-ip-prec(1))
Example:
qos-policy-out=add-class(sub, (class-default, data),
shape(400),set-ip-prec(1))
Step 3
Modifying Service Policy through CoA : Examples
//Policy-map configuration before CoA
policy-map __sub_5e311c4f_child1
class voip
priority level 1
police rate 10000 kbps burst 8 kbytes
!
!
class video
priority level 1
!
!
class data
shape average 80000 kbps
!
class class-default
!
end-policy-map
!
policy-map __sub_5e311c4f
class class-default
service-policy __sub_5e311c4f_child1
!
end-policy-map
!
//Modifying Service Policy through CoA
qos-policy-out=remove-class(sub, (class-default, voip))
qos-policy-out=add-class(sub, (class-default, video), bw-rpct(50), pri-level(2))
qos-policy-out=add-class(sub, (class-default, data), shape(400),set-ip-prec(1))
Release 4.3.x
188 OL-28375-03
//Policy-map configuration after CoA looks like:
policy-map __sub_ffffffec1a37f_child1
class video
priority level 2
bandwidth percent 50
!
!
class data
set precedence 1
!
class class-default
!
end-policy-map
!
policy-map __sub_ffffffec1a37f
class class-default
service-policy __sub_ffffffec1a37f_child1
!
end-policy-map
!
QoS Accounting
The QoS overhead accounting feature enables BNGto account for various encapsulation types when applying
QoS to packets. The ATM overhead accounting enables the BNG to account for the ATM encapsulation on
the subscriber line. It also accounts for the overhead added by cell segmentation. This accounting enables the
service provider to prevent overruns on the subscriber line and ensures that the BNG executes QoS features
on the actual bandwidth allocated to the subscriber traffic. The ATM overhead encapsulation details are listed
in this table.
Table 4: ATM Overhead Encapsulation Details
ALE Tags (RFC 4679) DSLAM to CPE
Encapsulation
Encapsulation2 Encapsulation1 Data Link Overhead (in bytes) CLI Option
PPPoA LLC (1) N/A AAL5 12 snap-pppoa
PPPoA Null (2) N/A AAL5 10 mux-pppoa
IPoA LLC (3) Untagged Ethernet AAL5 18 snap-1483routed
IPoA NULL (4) Untagged Ethernet AAL5 8 mux-1483routed
Ethernet over AAL5
LLC without FCS
(6)
Untagged Ethernet AAL5 28 snap-rbe
Ethernet over AAL5
LLC without FCS
(6)
Single-Tagged
Ethernet
AAL5 32 snap-dot1q-rbe
4.3.x
OL-28375-03 189
QoS Accounting
ALE Tags (RFC 4679) DSLAM to CPE
Encapsulation
Encapsulation2 Encapsulation1 Data Link Overhead (in bytes) CLI Option
Ethernet over AAL5
Null without FCS
(8)
Untagged Ethernet AAL5 24 mux-rbe
Ethernet over AAL5
Null without FCS
(8)
Single-Tagged
Ethernet
AAL5 28 mux-dot1q-rbe
To enable QoS overhead accounting, see Configuring QoS Accounting, on page 190.
Configuring QoS Accounting
Perform this task to enable QoS Layer2 overhead accounting.
SUMMARY STEPS
1. configure
2. dynamic-template
3. type [ppp|ip-subscriber|service]name
4. qos-account [ AAL5| user-defined ] [ mux-1483routed | mux-dot1q-rbe | mux-pppoa | mux-rbe |
snap-1483routed | snap-dot1q-rbe | snap-ppoa | snap-rbe ]
5. exit
DETAILED STEPS
Example:
Step 1
Enters dynamic template configuration mode. dynamic-template
Example:
Step 2
Specifies the type of dynamic template that
needs to be applied. Three type are:
type [ppp|ip-subscriber|service]name
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template)# type ppp p1
Step 3
PPP
IP-subscriber
Release 4.3.x
190 OL-28375-03
Configuring QoS Accounting
Service
Defines the L2 QoS overhead accounting.
Various keywords such as mux-1483routed,
qos-account [ AAL5| user-defined ] [ mux-1483routed | mux-dot1q-rbe
| mux-pppoa | mux-rbe | snap-1483routed | snap-dot1q-rbe | snap-ppoa |
snap-rbe ]
Step 4
snap-rbe define different available
encapsulations between the DSLAM and
CPE.
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template-type)# qos-account
AAL5 snap-rbe
For details about keywords, see Table 4:
ATM Overhead Encapsulation Details, on
page 189.
Exits from the current mode. exit
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template-type)# exit
Step 5
commitSaves the configuration changes,
and remains within the configuration session.
endPrompts user to take one of these
actions:
YesSaves configuration changes and
exits the configuration session.
NoExits the configuration session
changes.
mode, without committing the
Configuring QoS Accounting: An example
configure
dynamic-template type ppp p1
qos account AAL5 mux-1483routed
service-policy input input_1
end
Support for Shared Policy Instance
Shared Policy Instance (SPI) allows allocation of a single set of QoS resources among groups of BNG
sub-interfaces and bundle sub-interfaces, and shares them across a group of sub-interfaces, multiple Ethernet
flow points (EFPs), or bundle interfaces.
4.3.x
OL-28375-03 191
Support for Shared Policy Instance
Using SPI, a single instance of QoS policy can be shared across multiple sub-interfaces, allowing for aggregate
shaping of the sub-interfaces to one rate. All sub-interfaces that share the instance of a QoS policy must belong
to the same physical interface. The number of sub-interfaces sharing the QoS policy instance can range from
2 to the maximum number of sub-interfaces on the port.
For bundle interfaces, hardware resources are replicated per bundle member. All sub-interfaces that use a
common shared policy instance and are configured on a Link Aggregation Control Protocol (LAG) bundle
must be load-balanced to the same member link.
When a policy is configured on a bundle EFP, one instance of the policy is configured on each of the bundle
member links. When using SPI across multiple bundle EFPs of the same bundle, one shared instance of the
policy is configured on each of the bundle member links. By default, the bundle load balancing algorithm
uses hashing to distribute the traffic (that needs to be sent out of the bundle EFPs) among its bundle members.
The traffic for single or multiple EFPs can get distributed among multiple bundle members. If multiple EFPs
have traffic that needs to be shaped or policed together using SPI, the bundle load balancing has to be configured
to select the same bundle member (hash-select) for traffic to all the EFPs that belong the same shared instance
of the policy. This ensures that traffic going out on all the EFPs with same shared instance of the policy use
the same policer or shaper Instance.
BNG configures a complete hierarchical policy-map that includes parent and child policies. Optionally, the
SPI name can be defined and attached to the appropriate dynamic template or downloaded from RADIUS, in
this manner:
Policy configured through a CLI and applied through a dynamic-template
Policy configured through a CLI and applied through RADIUS
Restrictions
These restrictions apply to the usage of shared policy instance:
SPI is not supported for subscribers on non-bundle interfaces.
SPI is not supported for Parameterized QoS (PQoS). In a PQoS configuration, if there exists a SPI name,
then it is ignored.
SPI modified through CoA is not supported on subscribers.
Configuring a Policy with SPI in the Input or Output Direction Using Dynamic
Template
Perform this task to configure a policy with shared policy instance in the input and output direction using
dynamic template.
Release 4.3.x
192 OL-28375-03
Configuring a Policy with SPI in the Input or Output Direction Using Dynamic Template
SUMMARY STEPS
1. configure
2. policy-map policy_map_name
3. class {class_name | class-default | } [type qos]
4. service-policy service_policy_name
7. class {class_name | class-default | } [type qos]
8. police rate value
10. dynamic-template type ipsubscriber dynamic_template_name
11. service-policy {input |output}policy_map_name [shared-policy-instance instance_name]
12. service-policy {input |output}policy_map_name [shared-policy-instance instance_name]
DETAILED STEPS
Example:
Step 1
Creates or modifies a policy map that can be attached to one or
more interfaces to specify a service policy, and enters the
policy-map configuration submode.
policy-map policy_map_name
Example:
policy1
Step 2
Specifies the name of the class whose policy you want to create
or change and enters the policy map class configuration
class {class_name | class-default | } [type qos]
Example:
class-default
Step 3
submode. This example configures a traffic policy for the default
class of the traffic policy policy1. The default class is named
class-default.
Attaches a policy map to an input or output interface. service-policy service_policy_name
Example:
service-policy policy1_child
Step 4
4.3.x
OL-28375-03 193
Example:
policy1_child
Step 6
Specifies the name of the class whose policy you want to create
or change and enters the policy map class configuration
class {class_name | class-default | } [type qos]
Example:
class-default
Step 7
submode. This example configures a traffic policy for the default
class of the traffic policy policy1. The default class is named
class-default.
Configures traffic policing and enters policy map police
configuration mode. The value represents the committed
information rate and ranges from 1 to 4294967295.
police rate value
Example:
RP/0/RSP0/CPU0:router(config-pmap-c)# police
rate 1024
Step 8
Creates a dynamic template of type ipsubscriber. dynamic-template type ipsubscriber
dynamic_template_name
Step 10
Example:
type ppp PTA_TEMPLATE_1
Release 4.3.x
194 OL-28375-03
Attaches a policy map to an input or output interface to be used
as the service policy for that interface. In this example, the traffic
policy evaluates all traffic entering into that interface.
service-policy {input |output}policy_map_name
[shared-policy-instance instance_name]
Example:
input policy1 shared-policy-instance spi_1
Step 11
Attaches a policy map to an input or output interface to be used
as the service policy for that interface. In this example, the traffic
policy evaluates all traffic leaving that interface.
service-policy {input |output}policy_map_name
[shared-policy-instance instance_name]
Example:
output policy1 shared-policy-instance spi_2
Step 12
Configuring a Policy with SPI in the Input or Output Direction Using Dynamic Template: Example
configure
policy-map policy1
class class-default
!!
policy-map policy1_child
class class-default
police rate 1024 kbps
!!
dynamic-template
type ppp PTA_TEMPLATE_1
service-policy input policy1 shared-policy-instance spi_1
service-policy output policy1 shared-policy-instance spi_2
commit
4.3.x
OL-28375-03 195
Configuring a Policy with SPI in the Input or Output Direction Using RADIUS
Perform this task to configure a policy with shared policy instance in the input or output direction using
RADIUS.
SUMMARY STEPS
1. configure
3. class {class_name | class-default} [type qos]
4. service-policy service_policy_name
7. class {class_name | class-default} [type qos]
8. police rate value
DETAILED STEPS
Example:
Step 1
Example:
policy1
Step 2
Specifies the name of the class whose policy you want to create or
change and enters the policy map class configuration submode.
class {class_name | class-default} [type qos]
Example:
class-default
Step 3
This example configures a traffic policy for the default class of the
traffic policy policy1. The default class is named class-default.
Attaches a policy map to an input or output interface. service-policy service_policy_name
Example:
Step 4
Release 4.3.x
196 OL-28375-03
session.
Example:
policy1_child
Step 6
Specifies the name of the class whose policy you want to create or
change and enters the policy map class configuration submode.
class {class_name | class-default} [type qos]
Example:
class-default
Step 7
This example configures a traffic policy for the default class of the
traffic policy policy1. The default class is named class-default.
Configures traffic policing and enters policy map police
configuration mode. The value represents the committed information
rate and ranges from 1 to 4294967295.
police rate value
Example:
RP/0/RSP0/CPU0:router(config-pmap-c)# police
rate 1024
Step 8
session.
Configuring a Policy with SPI in the Input or Output Direction Using RADIUS: Example
configure
policy-map policy1
class class-default
!!
policy-map policy1_child
4.3.x
OL-28375-03 197
class class-default
police rate 1024 kbps
commit
!!
//In the USER file in RADIUS
RoadRunner_P1@Chasing1 Cleartext-Password := "LooneyTunes_P1"
cisco-avpair += "sub-qos-policy-in=policy1 shared-policy-instance spi_1",
cisco-avpair += "sub-qos-policy-out=policy1 shared-policy-instance spi_2",
Framed-Protocol += PPP,
Service-Type += Framed-User,
Fall-Through = no
What to Do Next
Run these steps in the USER file in RADIUS:
RoadRunner_P1@Chasing1 Cleartext-Password := "LooneyTunes_P1"
cisco-avpair += "sub-qos-policy-in=policy1 shared-policy-instance spi_1",
cisco-avpair += "sub-qos-policy-out=policy1 shared-policy-instance spi_2",
Framed-Protocol += PPP,
Service-Type += Framed-User,
Fall-Through = no
Merging QoS Policy-maps
Multiple QoS policies, applied through multiple dynamic templates, can be merged and implemented on a
single subscriber. The order in which the policies are merged is important, and is determined by the value of
the sequence number configured in the dynamic template. A policy is deployed using a policy-map. A new
optional merge keyword is provided with the service-policy command under dynamic template submode to
allow for the merging of policy-maps applied through multiple dynamic templates.
When more than two policy-maps are to be merged, two policy-maps are first merged together to create a
merged policy-map. Then, a third policy-map is merged with the first merged policy-map. This continues till
all policy-maps that are to be merged are merged together. For example, let's say that policy-maps p1, p2, p3,
p4 are to be merged in that order; p1 and p2 are merged first (using the rules listed below). Next, p3 is merged
with the <p1-p2> merged policy-map. Finally, p4 is merged with the <p1-p2-p3> merged policy-map, giving
the final merged policy-map.
The rules for merging two policy-maps are:
A merged policy-map can be created by appending the classes of the second policy-map to the classes
of the first policy-map (except for the default class).
If the same class (except for the default class) is configured under both the policies, the instance of that
class (including all actions configured under it) in the second policy is ignored.
If the default class under the first policy contains any actions other than any child policy actions, then
that default class is added to the end of the merged policy. If it contains any child policy actions, then
the default class from the second policy is added at the end of the merged policy.
If a child policy is configured under the default class of both policies, the two child policies are merged
using the rules above. The merged child policy is then applied as the child policy under the default class
of the merged parent policy.
If a child policy is configured under the default class of either the first or second policy (but not both),
then it is applied (as it is) as the child policy under the default class of the merged policy. Child policies
under classes other than the default class are never merged together.
Release 4.3.x
198 OL-28375-03
Merging QoS Policy-maps
If the sequence numbers of two policies to be merged are configured to be the same, the order in which
they are merged with respect to each other is random, and may change after the process restarts. Such
configurations must be avoided.
Note
Enabling Policy-maps Merge
Performthis task to enable merging of multiple QoS policy-maps applied through multiple dynamic templates.
SUMMARY STEPS
1. configure
2. dynamic-template
3. type service dynamic-template-name
4. service-policy {input | output | type} service-policy_name [acct-stats] [merge seq_num]
DETAILED STEPS
Example:
Step 1
Example:
Step 2
Creates a dynamic-template with a user-defined name
for a service.
type service dynamic-template-name
Example:
service s1
Step 3
Associates a service-policy to the dynamic template, and
enables merging of multiple QoS policies.
service-policy {input | output | type} service-policy_name
[acct-stats] [merge seq_num]
Example:
service-policy input QoS1 merge 10
service-policy output QoS2 merge 20
Step 4
4.3.x
OL-28375-03 199
Enabling Policy-maps Merge: Examples
dynamic-template type service default-service
service-policy input default-policy-in merge 100
service-policy output default-policy-out merge 100
!
dynamic-template type service voip-service
service-policy input voip-policy-in merge 20
service-policy output voip-policy-out merge 30
!
dynamic-template type service vod-service
service-policy input vod-policy-in merge 30
service-policy output vod-policy-out merge 50
!
dynamic-template type service turbo-button-service
service-policy input turbo-button-policy-in merge 10
service-policy output turbo-button-policy-out merge 40
!
end
\\the following configuration explains the merging behavior of egress qos policies
policy-map type qos default-policy-out
class class-default
shape average 2 mbps
bandwidth 512 kbps
service-policy default-policy-child-out
!
end-policy-map
policy-map type qos default-policy-child-out
class critical-data
set cos 3
queue-limit 500 ms
!
class best-effort-data
shape average percent 50
random-detect 100 ms 200 ms
set cos 5
!
class class-default
set cos 7
!
end-policy-map
policy-map type qos voip-policy-out
class class-default
service-policy voip-policy-child-out
Release 4.3.x
200 OL-28375-03
!
end-policy-map
policy-map type qos voip-policy-child-out
class voip-control
priority level 1
set cos 2
!
class voip-data
priority level 2
set cos 2
!
class class-default
!
end-policy-map
policy-map type qos vod-policy-out
class class-default
service-policy vod-policy-child-out
!
end-policy-map
policy-map type qos vod-policy-child-out
class vod-control
priority level 1
set cos 1
!
class vod-data
priority level 2
queue-limit 100 ms
!
class class-default
!
end-policy-map
policy-map type qos turbo-button-policy-out
class class-default
bandwidth 2 mpbs
!
end-policy-map
\\after the default and voip services are enabled on a subscriber session
policy-map type qos <merged-policy-1> !! Name is generated internally. This is just an
example.
class class-default
bandwidth 512 kbps
service-policy <merged-child-policy-1>
!
end-policy-map
policy-map type qos <merged-child-policy-1>
class voip-control
priority level 1
set cos 2
!
class voip-data
priority level 2
set cos 2
!
class critical-data
set cos 3
queue-limit 500 ms
!
4.3.x
OL-28375-03 201
set cos 5
!
class class-default
set cos 7
!
end-policy-map
\\after the turbo-button service is enabled
policy-map type qos <merged-policy-2>
class class-default
bandwidth 2 mpbs
service-policy <merged-child-policy-1> !! <merged-child-policy-1> is the same as before
since the
!! the turbo-button-policy-out does not have
any child policy
!! to be merged.
!
\\after the vod service is enabled
policy-map type qos <merged-policy-3>
class class-default
bandwidth 2 mbps
service-policy <merged-child-policy-2>
!
end-policy-map
policy-map type qos <merged-child-policy-1>
class voip-control
priority level 1
set cos 2
!
class voip-data
priority level 2
set cos 2
!
class vod-control
priority level 1
set cos 1
!
class vod-data
priority level 2
queue-limit 100 ms
!
class critical-data
set cos 3
queue-limit 500 ms
!
set cos 5
!
class class-default
set cos 7
!
end-policy-map
QoS Features Supported on BNG
BNG supports these QoS features:
Release 4.3.x
202 OL-28375-03
Policing and Queuing Support
BNGprovides ingress and egress traffic policers. BNGalso supports pre-existing traffic policing mechanisms
per subscriber session. 1R2Cand 2R3Cpolicers with marking actions is supported at parent-level in subscriber
policies. Only absolute police rates are supported at the parent-level of subscriber policies. 1R2C and 2R3C
policers with marking actions are supported at the child-level in subscriber policies. Both absolute and
percentage based police rates are supported at child-level of subscriber policies.
BNG supports traffic shaping at the physical port level, at the subscriber session level, at the class level, and
at the VLANlevel only in egress direction. The systemsupports all pre-existing queuing actions for subscriber
sessions. The configuration of minimum-bandwidth at the parent-level in subscriber policies is blocked. If
subscriber policies do not have a queuing action, the traffic on those subscribers is still subjected to S-VLAN
shaping and the traffic goes out through S-VLAN policy queues if those are present; if not, the traffic goes
through the interface default-queue. The shaping or bandwidth-remaining queuing action is mandatory in flat
S-VLAN policies. Only absolute shape rates is supported in S-VLAN flat policies and the parent-level of
subscriber policies. However, only shaping and bandwidth-remaining queuing actions are supported in the
parent-level of subscriber policies and all queuing actions are supported in the child-level of subscriber policies.
These additional queuing features are supported in egress policies applied on subscribers:
A policy can have 1 P1, 1 P2, 1 P3 and 5 normal priority queues.
A policy can have 1 P1, 2 P2 and 5 normal priority queues. P1 and P3 queues can be shared by multiple
classes whereas P2 queues are never shared.
Default Marking
BNG supports all pre-existing classification and marking options supported for L3 interfaces for use with
subscriber sessions. BNG also supports L3 marking to L2 marking mapping. BNG also supports ToS to CoS
mapping at LAC for downstream PPPoE frames and provides mechanisms to mark 802.1p and IP TOS fields.
The systemallows flexible IP TOS marking for L2TP packets based on ingress subscriber qos policy. Marking
is supported at the parent-level in subscriber policies and at the child-level in subscriber policies.
QoS Policy Modification
BNG supports in-service QOS policy-modification. Modification of subscriber-policy (through Radius),
S-VLAN policy (through CLI) and port sub-rate policy (through CLI) are also supported.
L2 Encapsulation
For PPPoE subscribers, the L2 encapsulation size used in QoS rate calculations must be adjustable based on
the last mile encapsulation (DSLAM to subscriber home) signaled in the PPPoE tags.
Classification
The BNG supports all pre-existing classification and marking options supported for L3 interfaces for use with
subscriber sessions. BNG also supports ingress classification based on 802.1P values for single and double
tagged COS, classification based on DSCP in either direction, classification based on L3/L4 ACLs in either
direction, and classification of L2TPv2 traffic based on the outer DSCP marking.
The classification of an incoming L2TP packet on the ingress core side interface is always based on the outer
IP fields even if the packet arrives with an MPLS tag stack.
4.3.x
OL-28375-03 203
Policy Inheritance
This table is relevant for egress direction only, as in ingress direction sub-rate policy and S-VLAN policy is
not supported:
Subscriber S-VLAN Port
Subscriber policy , if present, is
executed first; then, traffic is
subjected to port-shaper.
No policy is configured.
Inheritance limited to traffic getting
shaped by port sub-rate policy.
This is done irrespective of whether
a policy is configured on the
S-VLAN, or not.
Sub-rate policy
Subscriber policy is executed first,
if present, and then, S-VLAN
policy is executed. Finally traffic
is subjected to port-shaper.
Policy is configured. Inheritance
limited to traffic that gets shaped
by port sub-rate policy. This is
done irrespective of whether a
policy is configured on the
S-VLAN or not.
Sub-rate policy
Policy configuration is blocked and
port policy inherited through the
S-VLAN.
Policy configuration is blocked and
port policy is inherited.
HQoS or policy with more than
class-default
Subscriber policy is executed first,
if present, and then S-VLANpolicy
is executed.
Policy is configured. No policy configured
Subscriber with No QoS
When QoS is not configured on a subscriber, the parent S-VLAN, or on the port, subscriber traffic goes out
using the default-queue of its parents physical port.
The subscriber is subjected to the S-VLAN policy and goes out using S-VLAN policy queues, if those
are present. If the S-VLAN policy does not have its own queues, then all the S-VLAN traffic, including
the subscriber's, goes out through the default queue of the physical interface.
The subscriber is subject to a port policy, but no S-VLAN policy. Similar to the S-VLAN case, the
subscriber traffic is subject to it and uses its queues.
If a non-port-shaper policy is applied on the port, the application of policy on S-VLAN and subscriber
is blocked. In such a scenario, subscriber traffic is subjected to the policy applied on the port.
Control Packet Handling
BNG provides priority treatment in handling PPP Link Control Protocol (LCP) packets. The control packets
are handled in high priority without the need of user configuration, and these packets are not subjected to QoS
policies that are applied on both ingress and egress of the interface. In the case of LAC upstream direction,
if user wants a trusted COS value, then a PPP command is provided to impose the core-side header based on
the set trusted-COS. Thus, this ensures the priority treatment of these control packets in the network.
Release 4.3.x
204 OL-28375-03
S-VLAN Shaping and Statistics
In the egress direction, the BNG supports the ability to have policies at three different levels: the subscriber
interface level, the stacked virtual local area network (S-VLAN), and at the port level. The egress S-VLAN
and port-level policies are applied through CLI directly at the interface level. For applying a QoS policy on
S-VLAN, see Configuring Policy on S-VLAN, on page 207
The subscriber policy can only be applied through a dynamic template or via RADIUS. The egress subscriber
policy can be a two-level policy. The S-VLAN and port-level policies can only be flat policies, with only the
class default, with the only action being a shaped rate. Essentially it provides a means to constrain the S-VLAN
or port to a maximum rate via shaping.
In the ingress direction, the traffic is only subject to the subscriber input policy where the subscriber policies
are applied through RADIUS or dynamic-template.
The traffic through the S-VLAN includes traffic to many subscribers that may have already been shaped by
the subscriber policies. Providing statistics on that S-VLAN shaper is important in order to monitor whether
it is reaching the maximum capacity. Unlike the subscriber QoS policies, the HW does not have the ability
to directly track the usage or transmitted packets/bytes through this S-VLANshaper. So unlike other statistics,
the BNG provides the S-VLAN QoS policy-related statistics by aggregating the statistics of the underlying
subscriber policies. The statistics are displayed via show commands (and MIBs as appropriate) consistent
with all other interface types.
S-VLAN supports these conditions:
Modification of QoS rates.
Modification of S-VLAN policy to change number of levels in the policy is rejected.
Modification of two-level S-VLAN policy to add or remove child-level classes is rejected.
Modification of classification criteria in child-level classes, in two-level policy, is rejected.
Addition or removal of actions, in both two-level and flat policy, is rejected.
QoS Attachment Points
This table lists the QoS attachment points, and modes for definition and application.
Type of Policy Application Definition QoS Attachment Point
Flat class-default
only
CLI/XML CLI/XML Port (sub-rate policy)
Flat class-default
only. 2 level, with
parent class-default
only and child any
classification.
CLI/XML CLI/XML S-VLAN
2 level, with parent
class-default only and
child any
classification.
Dynamic-Template CLI/XML Subscriber
4.3.x
OL-28375-03 205
Type of Policy Application Definition QoS Attachment Point
child any
classification.
RADIUS CLI/XML Subscriber
child any
classification.
RADIUS RADIUS (parameterized QoS) Subscriber
Un-supported configurations will not be blocked. In S-VLANpolicies and subscriber policies, any configuration
other than the ones listed in these tables will be blocked:
Table 5: Supported Configuration in Ingress Direction
Rates Action Classification
Absolute only police, marking Class-default only Subscriber Parent Level
Policy
Absolute and percent police, marking Any, with baseline
restrictions
Subscriber Child Level
Policy
Table 6: Supported Configurations in Egress Direction
Rates Action Classification
Absolute only Any, with mandatory
shape action
Class-default only S-VLAN Flat Policy
Absolute only Any, with mandatory
shape action
Class-default only S-VLAN Parent Level
Policy
Absolute and percent Any Any, with baseline
restrictions
S-VLAN Child Level
Policy
Absolute only shape, bandwidth
remaining, police,
marking
Class-default only Subscriber Parent Level
Policy
Absolute and percent Any Any, with baseline
restrictions
Subscriber Child Level
Policy
VLAN Policy on Access Interface
BNG supports ingress and egress VLAN policies on an access-interface. Unlike as in the case of S-VLAN
(subscriber-parent) policy, the access-interface VLANpolicy is not inherited by the session policy. The VLAN
Release 4.3.x
206 OL-28375-03
VLAN Policy on Access Interface
policy does not provide reference bandwidth to session policies. The VLAN policy statistics does not include
session policy statistics. Only the access-interface traffic is subjected to the VLAN policy.
For details, see Configuring VLAN Policy on an Access Interface, on page 208.
This table summarizes the support for VLAN and S-VLAN policies in ingress and egress directions:
S-VLAN policy(with
subscriber-parent keyword)
V-LAN policy (without
subscriber-parent keyword)
Policy Direction
Not supported Supported Ingress
Supported Supported Egress
Restrictions
These restrictions apply to the VLANpolicy on the access-interface, when used without the subscriber-parent
keyword:
The VLAN policy needs to be attached to the access-interfaces, before bringing up the sessions with
QoS policies.
The restrictions specified for the in-place modification of S-VLAN policy, are applicable to VLAN
policy as well. For instance, the in-place modification for the VLAN policy supports only rate-changes.
This restriction also applies in adding a policer or shaper and in changing the policy-map to include
more classes.
Configuring Policy on S-VLAN
Perform this task to apply a QoS policy on a S-VLAN.
Note S-VLAN policy has to be provisioned before any policies are installed on subscribers.
Application of S-VLAN policy is rejected, if policies are already installed on subscribers.
Removal of S-VLAN policy is rejected, if subscriber policies are present under that S-VLAN.
SUMMARY STEPS
1. configure
2. interface type
3. service-policy output name subscriber-parent
4.3.x
OL-28375-03 207
Configuring Policy on S-VLAN
DETAILED STEPS
Example:
Step 1
Configures the subscribers on the Bundle-Ether access interface. interface type
Example:
Bundle-Ether1.1
Step 2
Configures the s-vlan policy with the subscriber-parent keyword. service-policy output name subscriber-parent
Example:
RP/0/RSP0/CPU0:router(config-if)#
Step 3
service-policy output svlan
subscriber-parent
session.
Configuring Policy on S-VLAN: An example
configure
service-policy output svlan_pmap subscriber-parent
end
!
Configuring VLAN Policy on an Access Interface
Perform this task to apply an ingress and egress QoS VLAN policy on an access interface.
Release 4.3.x
208 OL-28375-03
SUMMARY STEPS
1. configure
2. interface type
3. service-policy input service-policy-name
4. service-policy output service-policy-name
DETAILED STEPS
Example:
Step 1
Configures subscribers on the Bundle-Ether access interface. interface type
Example:
Bundle-Ether18.203
Step 2
Configures the ingress VLANQoS policy on the access-interface. service-policy input service-policy-name
Example:
RP/0/RSP0/CPU0:router(config-subif)#
service-policy input mark
Step 3
Configures the egress VLAN QoS policy on the access-interface. service-policy output service-policy-name
Example:
RP/0/RSP0/CPU0:router(config-subif)#
service-policy output metering
Step 4
4.3.x
OL-28375-03 209
Configuring Ingress and Egress VLAN Policies on an Access Interface: Example
//Attaching Ingress and Egress VLAN Policies on an Access Interface
configure
service-policy input INGRESS_MARKING_POLICING_POLICY
service-policy output VLAN_POLICY
end
!
//Attaching Ingress VLAN Policy and Egress S-VLAN Policies on an Access Interface
configure
service-policy input INGRESS_MARKING_POLICING_POLICY
service-policy output S_VLAN_POLICY subscriber-parent
end
!
These sections provide references related to implementing QoS.
MIBs
MIBs Link MIB
Link Description
Release 4.3.x
210 OL-28375-03
C HAP T E R 7
Configuring Subscriber Features
Subscriber features that are configured on BNG enable service providers to deploy certain specific
functionalities like restricting the use of certain network resources, allowing Law Enforcement Agencies
(LEAs) to conduct electronic surveillance, providing multicast services to the subscriber, and so on. The
subscriber features covered in this chapter are:
Excessive Punt Flow Trap, page 211
Access Control List and Access Control List-based Forwarding, page 215
Support for Lawful Intercept, page 218
TCP MSS Adjustment, page 226
Subscriber Session on Ambiguous VLANs , page 230
uRPF, page 232
Multicast Services, page 233
DAPS Support, page 245
HTTP Redirect Using PBR, page 256
Excessive Punt Flow Trap
The Excessive Punt Flow Trap feature attempts to identify and mitigate control packet traffic from remote
devices that send more than their allocated share of control packet traffic. A remote device can be a subscriber
device, a device on a VLAN interface, or a device identified by its source MAC address.
When remote devices send control packet traffic to the router, the control packets are punted and policed by
a local packet transport service (LPTS) queue to protect the router's CPU. If one device sends an excessive
rate of control packet traffic, the policer queue fills up, causing many packets to be dropped. If the rate from
one "bad actor" device greatly exceeds that of other devices, most of the other devices do not get any of their
control packets through to the router. The Excessive Punt Flow Trap feature addresses this situation.
4.3.x
OL-28375-03 211
Even when the Excessive Punt Flow Trap feature is not enabled, the "bad actors" can affect services for
only other devices; they cannot bring down the router.
Note
The Excessive Punt FlowTrap feature is supported on both subscriber interfaces, and non-subscriber interfaces
such as L2 and L3 VLAN sub-interfaces and bundle virtual interfaces (BVIs). If the source that floods the
punt queue with packets is a device with an interface handle, then all punts from that bad actor interface are
penalty policed. The default penalty rate, for each protocol, is 10 protocols per second (pps). Otherwise, if
the source is a device that does not have an interface handle, then all packets from this bad actor are dropped.
In the 4.2.x releases, the Excessive Punt FlowTrap feature was called as "Subscriber Control Plane Policing
(CoPP)" that only operated on subscriber interfaces.
Note
Functioning of Excessive Punt Flow Trap Feature
The Excessive Punt Flow Trap feature monitors control packet traffic arriving from physical interfaces,
sub-interfaces, BVI, and subscriber interfaces. It divides interfaces into two categories:
"Parent" interfaces, which can have other interfaces under them.
"Non-parent" interfaces, which have no interfaces under them.
Aphysical interface is always a parent interface because it has VLANsub-interfaces. ABVI is always a parent
interface because it is the "parent" of L2 interfaces. An L3 VLAN sub-interface can either be a parent or a
non-parent interface. If the VLANsub-interface is enabled for subscribers, then it is a parent interface, otherwise
it is a non-parent interface. A subscriber interface (IPoE or PPPoE) is always a non-parent interface.
When a flow is trapped, the Excessive Punt Flow Trap feature tries to identify the source of the flow. The
first thing it determines is from which interface the flow came. If this interface is not a "parent" interface,
then the feature assumes that it is the end-point source of the flowand penalty policing is applied. If the trapped
interface is a "parent" interface, then instead of penalizing the entire interface (which would penalize all the
interfaces under it), this feature takes the source MAC address of the bad flow and drops all packets from the
MAC address under the parent. Due to platform limitation, the penalty policer cannot be applied on a MAC
address; therefore all packets are dropped.
For more information about enabling the Excessive Punt Flow Trap feature, see Enabling Excessive Punt
Flow Trap Processing, on page 213.
The Excessive Punt Flow Trap feature monitors all punt traffic. There is no way to remove a particular
interface from the initial monitoring, nor can an interface be prevented from being flagged as bad if it is
the source of excessive flows.
Note
Bad actors are policed for each protocol. The protocols that are supported by the Excessive Punt Flow Trap
feature are Broadcast, Multicast, ARP, DHCP, PPP, PPPoE, ICMP, IGMP, L2TP and IP (covers many types
of L3 based punts, both IPv4 and IPv6). Each protocol has a static punt rate and a penalty rate. For example,
the sumtotal of all ICMP punts fromremote devices is policed at 1500 packets per second (pps) to the router's
CPU. If one remote device sends an excessive rate of ICMP traffic and is trapped, then ICMP traffic from
that bad actor is policed at 10 pps. The remaining (non-bad) remote devices continue to use the static 1500
pps queue for ICMP.
Release 4.3.x
212 OL-28375-03
Excessive Punt Flow Trap
The excessive rate required to cause an interface to get trapped has nothing to do with the static punt rate
(e.g. 1500 pps for ICMP). The excessive rate is a rate that is significantly higher than the current average
rate of other control packets being punted. The excessive rate is not a fixed rate, and is dependent on the
current overall punt packet activity.
Note
Once a bad actor is trapped, it is penalty policed on all its punted protocols (ARP, DHCP, PPP, etc.), irrespective
of the protocol that caused it to be identified as a bad actor. A penalty rate of 10 pps is sufficient to allow the
other protocols to function normally. However, if the bad actor is trapped by source MAC address, then all
its packets are dropped.
When an interface is trapped, it is placed in a "penalty box" for a period of time (a default of 15 minutes). At
the end of the penalty timeout, it is removed from penalty policing (or dropping). If there is still an excessive
rate of control packet traffic coming from the remote device, then the interface is trapped again.
Restrictions
These restrictions apply to implementing Excessive Punt Flow Trap feature:
This feature does not support interfaces on SIP-700 line cards and ASR 9000 Ethernet Line Card.
This feature is non-deterministic. In some cases, the Excessive Punt Flow Trap feature can give a false
positive, i.e. it could trap an interface that is sending legitimate punt traffic.
The Excessive Punt Flow Trap feature traps flows based on the relative rate of different flows; thus, the
behavior depends on the ambient punt rates. A flow that is significantly higher than other flows could
be trapped as a bad actor. Thus the feature is less sensitive when there are many flows, and more sensitive
when there are fewer flows present.
Sometimes control packet traffic can occur in bursts. The Excessive Punt Flow Trap has safeguards
against triggering on short bursts, but longer bursts could trigger a false positive trap.
Enabling Excessive Punt Flow Trap Processing
Perform this task to enable the Excessive Punt Flow Trap feature for both subscriber and non-subscriber
interfaces. The task also enables you to set the penalty policing rate and penalty timeout for a protocol.
SUMMARY STEPS
1. configure
2. lpts punt excessive-flow-trap subscriber-interfaces
3. lpts punt excessive-flow-trap non-subscriber-interfaces
4. lpts punt excessive-flow-trap penalty-rate protocol penalty_policer_rate
5. lpts punt excessive-flow-trap penalty-timeout protocol time
4.3.x
OL-28375-03 213
DETAILED STEPS
Example:
Step 1
Enables the Excessive Punt Flow Trap feature on subscriber
interfaces.
lpts punt excessive-flow-trap
subscriber-interfaces
Example:
RP/0/RSP0/CPU0:router(config)# lpts punt
excessive-flow-trap subscriber-interfaces
Step 2
Enables the Excessive Punt Flow Trap feature on non-subscriber
interfaces.
lpts punt excessive-flow-trap
non-subscriber-interfaces
Step 3
Example:
excessive-flow-trap
non-subscriber-interfaces
If both Step 2 and Step 3 configurations are applied, the
Excessive Punt Flow Trap feature is enabled for all
interfaces.
Note
Sets the penalty policing rate for a protocol. The penalty policer rate
is in packets-per-second (pps) and ranges from 2 to 100.
lpts punt excessive-flow-trap penalty-rate
protocol penalty_policer_rate
Step 4
Example:
excessive-flow-trap penalty-rate icmp 10
The penalty policing rate for a protocol consumes a policer
rate profile.
Note
Sets the penalty timeout value, which is a period of time that the
interface trapped is placed in the penalty box, for a protocol. The
lpts punt excessive-flow-trap penalty-timeout
protocol time
Step 5
penalty timeout value is in minutes and ranges from 1 to 1000. The
default penalty timeout value is 15 minutes.
Example:
excessive-flow-trap penalty-timeout igmp
10
session.
Release 4.3.x
214 OL-28375-03
Enabling Excessive Punt Flow Trap Processing: Examples
This is an example for enabling the Excessive Punt Flow Trap for subscriber interfaces, using the default
penalty timeout (15 minutes) and setting a penalty rate of 20 pps for PPP and PPPoE protocols.
configure
lpts punt excessive-flow-trap subscriber-interfaces
lpts punt excessive-flow-trap penalty-rate ppp 20
lpts punt excessive-flow-trap penalty-rate pppoe 20
end
!!
This is an example for enabling the Excessive Punt Flow Trap for non-subscriber interfaces, using the default
penalty rate (10 pps) and setting the ARP penalty timeout to 2 minutes.
configure
lpts punt excessive-flow-trap non-subscriber-interfaces
lpts punt excessive-flow-trap penalty-timeout arp 2
end
!!
Access Control List and Access Control List-based Forwarding
An Access Control List (ACL) is used to define access rights for a subscriber. It is also used for filtering
content, blocking access to various network resources, and so on.
Certain service providers need to route certain traffic be routed through specific paths, instead of using the
path computed by routing protocols. For example, a service provider may require that voice traffic traverse
through certain expensive routes, but data traffic to use the regular routing path. This is achieved by specifying
the next-hop address in the ACLconfiguration, which is then used for forwarding packet towards its destination.
This feature of using ACL for packet forwarding is called ACL-based Forwarding (ABF).
The ACL is defined through CLI or XML; however, it can be applied to a subscriber session either through
a dynamic-template, or through VSAs from RADIUS. Deploying ABF (using ACL) involves these stages:
Defining an ACL, see Configuring Access-Control Lists, on page 215.
Applying the ACL to an access-interface, see Activating ACL, on page 217.
Configuring Access-Control Lists
Perform this task to create an access control list. As an example, this access list is created to deploy ABF;
therefore, it defines the next hop address.
SUMMARY STEPS
1. configure
2. {ipv4 | ipv6} access-list access-list-name
3. sequence-number permit tcp any any
4. sequence-number permit {ipv4 | ipv6} host source_address nexthop source_address destination_address
4.3.x
OL-28375-03 215
Access Control List and Access Control List-based Forwarding
DETAILED STEPS
Example:
Step 1
Configures the access-list. {ipv4 | ipv6} access-list access-list-name
Example:
RP/0/RSP0/CPU0:router(config)# ipv4 access-list
foo_in
Step 2
or
RP/0/RSP0/CPU0:router(config)# ipv6 access-list
foo_in
Enters an access control list rule to tcp traffic. sequence-number permit tcp any any
Example:
RP/0/RSP0/CPU0:router(config)# 10 permit tcp any
any
Step 3
Specifies packets to forward on ipv4 protocol from source
IP address to destination IP address.
sequence-number permit {ipv4 | ipv6} host
source_address nexthop source_address
destination_address
Step 4
Repeat steps 1 to 4 to configure the foo_out
access-list.
Note
Example:
RP/0/RSP0/CPU0:router(config)# 10 permit ipv4
host 9.8.8.9 nexthop 6.6.6.6 7.7.7.7
or
RP/0/RSP0/CPU0:router(config)# 10 permit ipv6
host 192:2:1:9 nexthop 192:2:6:8
Release 4.3.x
216 OL-28375-03
Configuring Access-Control Lists
Configuring Access-Control Lists: Examples
//For IPv4
configure
ipv4 access-list foo_in
10 permit tcp any any
10 permit ipv4 host 9.8.8.9 nexthop 6.6.6.6 7.7.7.7
!
!
end
//For IPv6
configure
ipv6 access-list foo_in
10 permit tcp any any
10 permit ipv4 host 192:2:1:9 nexthop 192:2:6:8
!
!
end
Activating ACL
Perform this task to define a dynamic-template that is used to activate an access-control list.
SUMMARY STEPS
1. configure
2. dynamic-template
3. type{ipsubscriber |ppp |service} dynamic-template-name
4. {ipv4 | ipv6} access-group access-list-name ingress
DETAILED STEPS
Example:
Step 1
Example:
Step 2
Creates a service dynamic-template type. type{ipsubscriber |ppp |service} dynamic-template-name
Example:
service foo
Step 3
4.3.x
OL-28375-03 217
Activating ACL
Specifies access-control for the incoming packets. {ipv4 | ipv6} access-group access-list-name ingress Step 4
Example:
access-group foo_in ingress
Similarly, create another access-group for
the outgoing packets called foo_out.
Note
or
access-group foo_in ingress
Activating ACL: Examples
//For IPv4
configure
dynamic-template
type service foo
ipv4 access-group foo_in ingress
!
!
end
//For IPv6
configure
dynamic-template
type service foo
ipv6 access-group foo_in ingress
!
!
end
Support for Lawful Intercept
Lawful Intercept allows Law Enforcement Agencies (LEAs) to conduct electronic surveillance as authorized
by judicial or administrative order. Increasingly, legislation is being adopted and regulations are being enforced
that require service providers (SPs) and Internet service providers (ISPs) to implement their networks to
explicitly support authorized electronic surveillance. The types of SPs or ISPs that are subject to Lawful
Release 4.3.x
218 OL-28375-03
Support for Lawful Intercept
Intercept mandates vary greatly from country to country. Lawful Intercept compliance in the United States is
specified by the Communications Assistance for Law Enforcement Act (CALEA).
Cisco ASR 9000 Series Router supports the Cisco Service Independent Intercept (SII) architecture and
PacketCable
TM1
Lawful Intercept architecture. The Lawful Intercept components by themselves do not ensure
customer compliance with applicable regulations but rather provide tools that can be used by SPs and ISPs
to construct an Lawful Intercept compliant network.
BNG supports the Per-session Lawful Intercept and Radius-based Lawful Intercept for subscribers. Both,
per-session and radius-based lawful intercepts are executed on IPoE, PPPoE, and PPPoE LAC subscriber
sessions in BNG.
This guide does not address legal obligations for the implementation of lawful intercept. Service providers
are responsible for ensuring that network complies with applicable lawful intercept statutes and regulations.
It is recommended that legal advice be sought to determine obligations.
Caution
By default, Lawful Intercept is not a part of the Cisco IOS XR software. To enable Lawful Intercept, you
must install and activate the asr9k-li-px.pie.
For more information about Lawful Intercept-related router configuration, see Implementing Lawful
Intercept chapter in Cisco ASR 9000 Series Aggregation Services Router System Security Configuration
Guide.
Note
Per-session Lawful Intercept
Lawful interception of all Layer 2 or Layer 3 traffic on a specified subscriber interface, on both ingress as
well egress directions, and sending the replicated stream to mediation device, is called the per-session Lawful
Intercept. This Lawful Intercept implements IPv4, IPv6, and multicast traffic interception using the
Cisco-defined MIBs. By default, the SNMP-based Lawful Intercept feature is enabled on the Cisco ASR9000
Series Router, which allows you to configure the taps. For more information about disabling SNMP-based
Lawful Intercept, see Disabling SNMP-based Lawful Intercept, on page 220.
The subscriber session is identified by Account-session-ID, which acts as a key in identifying the specified
subscriber interface for the subscriber user, whose traffic is getting intercepted.
Lawful Intercept, in general, can be implemented using either SII architecture or PacketCable
TM
specifications.
The Cisco IOS-XRimplementation of SNMP-based Lawful Intercept is based on service-independent intercept
(SII) architecture. SNMPv3 authenticates data origin and ensures that the connection from Cisco ASR 9000
Series Router to the mediation device is secure. This ensures that unauthorized parties cannot forge an intercept
target.
1
PacketCable
TM
architecture addresses device interoperability and product compliance issues using the PacketCable
Specifications.
4.3.x
OL-28375-03 219
To implement lawful intercept, you must understand how the SNMP server functions. For this reason,
carefully review the information described in the module Implementing SNMP in Cisco ASR 9000 Series
Aggregation Services Router System Management Configuration Guide.
Lawful intercept must be explicitly disabled. It is automatically enabled on a provisioned router after
installing and activating the asr9k-li-px.pie. However, you should not disable LI if there is an active tap
in progress, because this deletes the tap.
Management plane must be configured to enable SNMPv3. Allows the management plane to accept SNMP
commands, so that the commands go to the interface (preferably, a loopback) on the router. This allows
the mediation device (MD) to communicate with a physical interface. For more information about
Management Plane Protection feature, see Configuring the Inband Management Plane Protection Feature,
on page 221 and for more information about enabling the mediation device, see Enabling the Mediation
Device to Intercept VoIP and Data Sessions, on page 221.
Note
Lawful Intercept MIBs
An external mediation device also known as collectors can create IPv4 or IPv6 address based TAPs using
IP-TAP-MIB. The SNMPv3 protocol is used to provision the mediation device (defined by CISCO-TAP2-MIB)
and the Taps(defined by CISCO-USER-CONNECTION-TAP-MIB). The Cisco ASR 9000 Series Router
supports a total of 511 concurrent taps that includes both SNMP and Radius.
Lawful intercept uses these MIBs for interception:
CISCO-TAP2-MIBUsed for lawful intercept processing. It contains SNMP management objects that
control lawful intercepts on a Cisco ASR 9000 Series Router. The mediation device uses the MIB to
configure and run lawful intercepts on targets sending traffic through the Cisco ASR9000 Series Router.
The CISCO-TAP2-MIB supports the SII feature and defines the provisioning of the mediation devices
and generic Taps. It primarily consists of the mediation device table and a stream table. The mediation
device table contains information about mediation devices with which the Cisco ASR9000 Series Router
communicates; for example, the device's address, the interfaces to send intercepted traffic over, and the
protocol to use to transmit the intercepted traffic. The stream table contains a list of generic Taps that
are provisioned by the MD table entries.
CISCO-USER-CONNECTION-TAP-MIBUsed for intercepting traffic for individual subscribers.
The MIB contains SNMP management objects to configure and execute wiretaps on individual user
connections on the Cisco ASR 9000 Series Router. This MIB contains information about the user
connections, each identified by a unique session ID. The CISCO-USER-CONNECTION-TAP-MIB
cannot be configured without configuring the CISCO-TAP2-MIB.
It is not possible to configure an SNMP tap and a Radius tap at the same time. Also, the same session
cannot be tapped more than once at a time.
Note
Disabling SNMP-based Lawful Intercept
Lawful Intercept is enabled by default on the Cisco ASR 9000 Series Router after installing and activating
the asr9k-li-px.pie.
Release 4.3.x
220 OL-28375-03
To disable Lawful Intercept, enter the lawful-intercept disable command in global configuration mode.
To re-enable it, use the no form of this command.
Disabling SNMP-based Lawful Intercept: An example
RP/0/RSP0/CPU0:router(config)# lawful-intercept disable
The lawful-intercept disable command is available only after installing and activating the asr9k-li-px.pie.
All SNMP-based taps are dropped when lawful intercept is disabled.
Note
Configuring the Inband Management Plane Protection Feature
If MPP was not earlier configured to work with another protocol, then ensure that the MPP feature is also not
configured to enable the SNMP server to communicate with the mediation device for lawful interception. In
such cases, MPP must be configured specifically as an inband interface to allow SNMP commands to be
accepted by the router, using a specified interface or all interfaces.
Ensure this task is performed, even if you have recently migrated to Cisco IOS XR Software from Cisco
IOS, and you had MPP configured for a given protocol.
Note
For lawful intercept, a loopback interface is often the choice for SNMP messages. If you choose this interface
type, you must include it in your inband management configuration.
Enabling the Mediation Device to Intercept VoIP and Data Sessions
These SNMP server configuration tasks enable the Cisco SII feature on a router running Cisco IOS XRSoftware
by allowing the MD to intercept VoIP or data sessions.
SUMMARY STEPS
1. configure
2. snmp-server view view-name ciscoTap2MIB included
3. snmp-server view view-name ciscoUserConnectionTapMIB included
4. snmp-server group group-name v3auth read view-name write view-name notify view-name
5. snmp-server host ip-address traps version 3 auth username udp-port port-number
6. snmp-server user mduser-id groupname v3 auth md5 md-password
8. show snmp users
9. show snmp group
10. show snmp view
4.3.x
OL-28375-03 221
DETAILED STEPS
Example:
Step 1
Creates or modifies a view record and includes the CISCO-TAP2-MIB
family in the view. The SNMP management objects in the
snmp-server view view-name ciscoTap2MIB
included
Step 2
CISCO-TAP2-MIB that controls lawful intercepts are included. This
Example:
RP/0//CPU0:router(config)# snmp-server
view TapName ciscoTap2MIB included
MIB is used by the mediation device to configure and run lawful
intercepts on targets sending traffic through the router.
Creates or modifies a view record and includes the
CISCO-USER-CONNECTION-TAP-MIBfamily, to manage the Cisco
snmp-server view view-name
ciscoUserConnectionTapMIB included
Step 3
intercept feature for user connections. This MIB is used along with the
CISCO-TAP2-MIB to intercept and filter user traffic.
Example:
view TapName ciscoUserConnectionTapMIB
included
Configures a newSNMP group that maps SNMP users to SNMP views.
This group must have read, write, and notify privileges for the SNMP
view.
snmp-server group group-name v3auth read
view-name write view-name notify view-name
Example:
Step 4
group TapGroup v3 auth read TapView write
TapView notify TapView
Specifies SNMP trap notifications, the version of SNMP to use, the
security level of the notifications, and the recipient (host) of the
notifications.
snmp-server host ip-address traps version 3
auth username udp-port port-number
Example:
Step 5
host 223.255.254.224 traps version 3 auth
bgreen udp-port 2555
Configures the MDuser as part of an SNMP group, using the v3 security
model and the HMAC MD5 algorithm, which you associate with the
MD password.
snmp-server user mduser-id groupname v3
auth md5 md-password
Example:
Step 6
The mduser-id and mdpassword must match that configured on
MD. Alternatively, these values must match those in use on the
router.
mduser-id TapGroup v3 auth md5 mdpassword
Passwords must be eight characters or longer to comply with
SNMPv3 security minimums.
Minimum Lawful Intercept security level is auth; The noauth
option will not work, as it indicates noAuthnoPriv security level.
Release 4.3.x
222 OL-28375-03
The Lawful Intercept security level must also match that of the
MD.
Choices other than MD5 are available on the router, but the MD
values must match.
Most MDs default to or support only MD5.
session.
Displays information about each SNMP username in the SNMP user
table.
show snmp users
Example:
RP/0//CPU0:router# show snmp users
Step 8
Displays information about each SNMP group on the network. show snmp group
Example:
RP/0//CPU0:router# show snmp group
Step 9
Displays information about the configured views, including the
associated MIB view family name, storage type, and status.
show snmp view
Example:
RP/0//CPU0:router# show snmp view
Step 10
Enabling the Mediation Device to Intercept VoIP and Data Sessions: An example
configure
snmp-server view TapName ciscoTap2MIB included
snmp-server view TapName ciscoUserConnectionTapMIB included
snmp-server group TapGroup v3 auth read TapView write TapView notify TapView
snmp-server host 223.255.254.224 traps version 3 auth bgreen udp-port 2555
snmp-server mduser-id TapGroup v3 auth md5 mdpassword
end
!
!
4.3.x
OL-28375-03 223
Radius-based Lawful Intercept
Radius-based Lawful Intercept feature provides mechanisms for interception of the BNG subscriber traffic
by using the RADIUS attributes. This is a preferred method over SNMP user-connection MIB, as SNMP-based
method prevents a session to be tapped until an IP address has been assigned to the session. In the Radius-based
LI mechanism, tapping is possible as soon as a session is established.
A RADIUS-based Lawful Intercept solution enables intercept requests to be sent (through Access-Accept
packets or Change of Authorization (CoA)-Request packets) to the network access server (NAS) or to the
Layer 2 Tunnel Protocol access concentrator (LAC) from the RADIUS server. All traffic data going to or
from a PPP or L2TP session is passed to a mediation device. Another advantage of RADIUS-based Lawful
Intercept solution is to set the tap with Access-Accept packets that allows all target traffic to be intercepted
simultaneously.
The RADIUS-based Lawful Intercept feature provides tap initiation support for these modes:
Access-Accept based Lawful Intercept for the new session
CoA based Lawful Intercept for existing session
By default, the Radius-based Lawful Intercept functionality is not enabled. For more information about
enabling Radius-based Lawful Intercept, see Enabling RADIUS-based Lawful Intercept, on page 224.
Note
Enabling RADIUS-based Lawful Intercept
Perform this task to enable the Radius-based Lawful Intercept feature.
SUMMARY STEPS
1. configure
2. aaa intercept
3. aaa server radius dynamic-author
4. port port_number
5. server-key [0|7] word
6. client hostname{ vrf vrf_name | server-key [0|7] word }
DETAILED STEPS
Example:
Step 1
Release 4.3.x
224 OL-28375-03
Enables the radius-based lawful intercept feature. aaa intercept Step 2
Example:
intercept
This command is available only after installing and
activating asr9k-li-px.pie.
When you disable aaa intercept, all radius-based taps are
removed from the Cisco ASR 9000 Series Router.
Note
Configures the lawful intercept as a AAA server and enters the
dynamic authorization local server configuration mode.
aaa server radius dynamic-author
Example:
RP/0/RSP0/CPU0:router(config)# aaa server
radius dynamic-author
Step 3
Specifies the RADIUS server port. The default port number is 1700. port port_number
Example:
RP/0/RSP0/CPU0:router(config-Dynamic
Author)# port 1600
Step 4
Specifies the encryption key shared with the RADIUS client. server-key [0|7] word
Example:
Author)# server-key cisco
Step 5
Specifies the client with which the AAA server will be
communicating.
client hostname{ vrf vrf_name | server-key [0|7]
word }
Step 6
Example:
Author)# client 3.0.0.28 vrf default
server-key cisco
You can configure the server key in a global mode and also
as a per client type key.
Note
session.
Enabling RADIUS-based Lawful Intercept: An example
configure
aaa intercept
aaa server radius dynamic-author
port 1600
4.3.x
OL-28375-03 225
server-key cisco
client 3.0.0.28 vrf default server-key cisco
end
!
!
What to Do Next
These attributes need to be present in the user profile to configure the Radius-based Lawful Intercept.
xyz_user1@domain.com Password == "cisco"
Cisco-avpair = "md-ip-addr=192.1.1.4",
Cisco-avpair += "md-port=203",
Cisco-avpair += "md-dscp=3",
Cisco-avpair += "intercept-id=abcd0003",
Cisco-avpair += "li-action=1"
TCP MSS Adjustment
The TCP MSS Adjustment feature allows the configuration of the maximum segment size (MSS) on transient
packets that traverse a Cisco ASR 9000 Series Router.
When dealing with PPPoE or L2TP cases, an additional header that the client initiating a TCP session may
not be aware of is added to the packet. This can result in lost packets, broken transmissions, or fragmentation
when packet sizes exceed the maximum transmission units (MTUs) due to the added headers.
Here is a sample scenario that shows how the TCP MSS adjust feature works:
Figure 9: Sample TCP MSS Adjust
In this example, the HTTP client sends to the HTTP server a TCP synchronize (SYN) packet that signals an
MSS value of 1300 (MTU) - 20 TCP - 20 IP header = 1260. On receiving it, the HTTP server acknowledges
it with a SYN ACK message. The HTTP client confirms the TCP session with a single acknowledgment and
opens up the TCP channel.
This is a sample scenario without PPPoE or L2TP. Note
When the HTTP server picks up a large file, it segments it into 1460 byte chunks (assuming that there are no
http headers for now). When the HTTP server sends the packet, the first Cisco ASR 9000 Series Router (on
the right) detects that the MTU is 576 downstream to the client and requires a 1300 byte packet to be
fragmented.
If the server sets the DF ("don't fragment") bit, then the packet is dropped. And, if the packet does not have
the DF bit set, then it gets fragmented, requiring the client to reassemble the packets. In digital subscriber line
(DSL) or fibre-to-the-home (FTTH) like access, a CPEmay block incoming fragments as a security mechanism,
causing this transmission to be lost.
In a typical scenario, having packets that are dropped causes partial downloads, an obstruction, or a delay in
displaying images in web pages. MSS adjust overcomes this scenario by intercepting the TCP SYN packet,
Release 4.3.x
226 OL-28375-03
TCP MSS Adjustment
reading the MSS option, and adjusting the value so that the server does not send packets larger than the
configured size (plus headers).
Note that the TCP MSS value is only adjusted downward. If the clients request an MSS value lower than the
configured value, then no action is taken.
In the case of PPPoE, an extra 8 bytes and in the case of L2TP, an extra 40 bytes is added to the packet. The
recommended MSS adjust values are 1452 for PPPoE, and 1420 for L2TP scenarios, assuming a minimum
MTU of 1500 end-to-end.
Separate unique global values for PTA and L2TP are supported, which once configured allows all future
sessions to beTCP MSS adjustment; however, the sessions already established will not be TCP adjusted. If
the global value is changed, then all new TCP subscriber sessions, will get the new global value.
For more information about configuring the TCP MSS value of packets, see Configuring the TCP MSS Value
of TCP Packets, on page 227.
To disable this on a session, you must first disable the global configuration, then delete the session and
recreate it.
Note
TCP encapsulated in both IPv4 and IPv6 are supported.
Restrictions
These restrictions are applicable for TCP MSS Adjustment:
Because the MSS is TCP-specific, the TCP MSS Adjustment feature is applicable only to (transit) TCP
packets and the UDP packets are unaffected.
TCP MSS Adjustment configuration affects only the PPPoE PTA and LAC sessions types. It does not
affect IP sessions or any non-BNG interfaces.
The MSS option must be the first option in the TCP header.
The router uses the MSS value that the user configures for checking TCP/IPV4 packets. When checking
TCP/IPv6 packets, the router automatically adjusts the configured MSS value down by 20 bytes to
account for the larger IPv6 header. For example, if the TCP MSS value is configured to 1450, then the
router adjusts the TCP MSS in an IPV4 packet down to 1450 and down to 1430 for an IPv6 packet.
Configuring the TCP MSS Value of TCP Packets
Perform this task to configure the TCP MSS value of TCP packets in order to prevent TCP sessions from
being dropped.
4.3.x
OL-28375-03 227
SUMMARY STEPS
1. configure
2. subscriber
3. pta tcp mss-adjust max-segment-size
5. configure
6. vpdn
7. l2tp tcp-mss-adjust max-segment-size
DETAILED STEPS
Example:
Step 1
Enables the subscriber configuration mode. subscriber
Example:
RP/0/RSP0/CPU0:router(config)# subscriber
Step 2
Sets the MSS value of TCP packets going through a Cisco ASR
9000 Series Router for a PTA subscriber. The TCP MSS Adjust
maximum segment size ranges from 1280 to 1536 (in bytes).
pta tcp mss-adjust max-segment-size
Example:
RP/0/RSP0/CPU0:router(config-subscriber)# pta
tcp mss-adjust 1300
Step 3
The value represents the global value for the PTA
sessions, when the feature is enabled it applies to all
sessions.
Note
Release 4.3.x
228 OL-28375-03
Example:
Step 5
Enables the vpdn configuration mode. vpdn
Example:
Step 6
Sets the MSS value of TCP packets going through a Cisco ASR
9000 Series Router for a LAC subscriber. The TCP MSS Adjust
maximum segment size ranges from 1280 to 1460 (in bytes).
l2tp tcp-mss-adjust max-segment-size
Example:
tcp-mss-adjust 1300
Step 7
Configuring the TCP MSS Value of TCP Packets: Examples
//Example for configuring the TCP MSS value of TCP packets for a PPPoE PTA subscriber
session:
configure
subscriber
pta tcp mss-adjust 1280
!!
// Example for configuring the TCP MSS value of TCP packets for a PPPoE LAC subscriber
session:
configure
vpdn
l2tp tcp-mss-adjust 1460
!!
4.3.x
OL-28375-03 229
Subscriber Session on Ambiguous VLANs
Ambiguous VLANenables you to create multiple subscriber sessions on a single access-interfaces. As a result,
it increases the scalability of the access-interface. An ambiguous VLAN is an L3 interface on which either a
VLAN ID range, or a group of individual VLAN IDs are specified. Instead of individually mapping each
subscriber to a VLAN, an ambiguous VLAN configuration performs the mapping for a group. Multiple
subscribers can be mapped on the ambiguous VLAN as long as they possess a unique MAC address. The
subscriber sessions created over ambiguous VLANs are identical to the ones created over regular VLANs,
and support all regular configurations such as policy-map, VRFs, QoS, access-control list, and so on.
For enabling IPoE subscriber session creation on an ambiguous VLAN, see Establishing Subscriber Session
on Ambiguous VLANs, on page 230.
Restriction
Ambiguous VLANs do not work with Unicast client.
Establishing Subscriber Session on Ambiguous VLANs
Perform this task to define an ambiguous VLAN and enable creation of IP subscriber session on it.
There is no DHCP-specific configuration required for ambiguous VLANs. Note
SUMMARY STEPS
1. configure
3. Use any of these commands to configure encapsulated ambiguous VLANs:
encapsulation ambiguous { dot1q | dot1ad } {any | vlan-range }
encapsulation ambiguous dot1q vlan-id second-dot1q { any | vlan-range }
encapsulation ambiguous dot1q any second-dot1q { any | vlan-id }
encapsulation ambiguous dot1ad vlan-id dot1q { any | vlan-range }
4. ipv4 | ipv6address source-ip-address destination-ip-address
7. initiator dhcp
Release 4.3.x
230 OL-28375-03
Subscriber Session on Ambiguous VLANs
DETAILED STEPS
Example:
Step 1
Configures the interface and enters the interface
configuration mode.
Example:
GigabitEthernet0/1/0/0.12
Step 2
Configures IEEE 802.1Q VLAN configuration. Use any of these commands to configure encapsulated ambiguous
VLANs:
Step 3
The vlan-range is given in comma-separated, or
hyphen-separated format, or a combination of both, as
shown in the examples.
encapsulation ambiguous { dot1q | dot1ad } {any |
vlan-range }
encapsulation ambiguous dot1q vlan-id second-dot1q {
any | vlan-range }
encapsulation ambiguous dot1q any second-dot1q { any
| vlan-id }
encapsulation ambiguous dot1ad vlan-id dot1q { any |
vlan-range }
Example:
ambiguous dot1q any
ambiguous dot1q 14 second-dot1q 100-200
ambiguous dot1q any second-dot1q any
ambiguous dot1ad 14 dot1q 100,200,300-400
Configures the IPv4 or IPv6 protocol address. ipv4 | ipv6address source-ip-address destination-ip-address
Example:
RP/0/RSP0/CPU0:router(config-if)# ipv4 address 2.1.12.1
Step 4
255.255.255.0
RP/0/RSP0/CPU0:router(config-if)# ipv6 address 1:2:3::4
128
Applies a policy-map to an access interface where the
policy-map was previously defined with the specified
PL1 policy_name.
Example:
control subscriber PL1
Step 5
4.3.x
OL-28375-03 231
Establishing Subscriber Session on Ambiguous VLANs
Enables l2-connected IPv4 IP subscriber. ipsubscriber ipv4 l2-connected
Example:
l2-connected
Step 6
Enables initiator DHCP on the IP subscriber. initiator dhcp
Example:
RP/0/RSP0/CPU0:router(config-if)# initiator dhcp
Step 7
Establishing Subscriber Session on Ambiguous VLANs: An example
configure
encapsulation ambiguous dot1q 14 second-dot1q any
ipv4 address 2.1.12.12 55.255.255.0
!
!
end
uRPF
Unicast Reverse Path Forwarding (uRPF) is a feature in BNGthat verifies whether the packets that are received
on a subscriber interface are sent froma valid subscriber. uRPF only applies to subscribers using an L3 service.
For PPPoE subscribers, the uRPF check ensures that the source address in the arriving packet matches the set
of addresses associated with the subscriber. The subscriber addresses are the IPCP assigned addresses, or any
framed routed assigned through RADIUS. PPPoE subscribers are identified by session ID and VLAN keys.
BNGperforms the uRPF check to ensure that the source IP address in the arriving packets matches the expected
session IDs and VLAN keys.
For IPoE subscribers, the subscriber addresses are the ones assigned through DHCP. IPoE subscribers are
identified by the incoming MAC address. The uRPF check ensures that the source IP address is the one
allocated by DHCP to the source MAC address.
Release 4.3.x
232 OL-28375-03
uRPF
uRPF is supported on both IPv4 and IPv6 subscribers and is enabled using a dynamic template. To define a
dynamic template for enabling uRPF, see Creating Dynamic Template for IPv4 or IPv6 Subscriber Session,
on page 84.
Multicast Services
Multicast services enable multiple subscribers to be recipients of a single transmission from one source. For
example, real-time audio and video conferencing makes good use of a multicast service. The multicast features
applied on the PPPoE interfaces of BNG includes:
Multicast Coexistence
On BNG, the multicast services coexist with regular unicast services. The multicast feature on BNG is the
same as the existing L3 multicast feature already supported on the Cisco ASR 9000 Series Routers. On BNG,
multicast is enabled on the trunk interfaces, and the VLANs created over physical interfaces and bundles.
Multicast co-existence works for PPPoEPTAsubscriber sessions. For more details on multicast implementation
on ASR9k, see Implementing Layer-3 Multicast Routing on Cisco IOS XR Software chapter in Cisco ASR 9000
Series Aggregation Services Router Multicast Configuration Guide.
To enable multicast function on BNG, see Enabling Address Family for the VRF, on page 233.
Enabling Address Family for the VRF
Perform this task to enable multicast functions for the required address family.
SUMMARY STEPS
1. configure
2. multicast-routing
3. vrf vrf_name
4. address-family ipv4
DETAILED STEPS
Example:
Step 1
Configures multicast-routing. multicast-routing
Example:
RP/0/RSP0/CPU0:router(config)# multicast
routing
Step 2
4.3.x
OL-28375-03 233
Multicast Services
Configures the vrf name. vrf vrf_name
Example:
RP/0/RSP0/CPU0:router(config)# vrf vrf1
Step 3
Enables the multicast functions in the ipv4 address family. address-family ipv4
Example:
address-family ipv4
Step 4
Enabling Address Family for the VRF: An example
multicast-routing
vrf vrf1
address-family ipv4
!
!
end
Multicast Replication
BNG supports the multicast packet replication on PPPoE interfaces. It also supports multicast forwarding on
subscriber interfaces, and transmission of multicast IP video content. When the multicast replication is enabled
for a subscriber, BNG performs IGMP statistics gathering for that subscriber, and has the ability to export
them. Multicast replication is supported on subscriber interfaces, which are configured in the passive mode.
HQoS Correlation
The Hierarchical quality of service (HQoS) correlation feature monitors every subscriber's multicast bandwidth
usage through IGMP reports received on each subscriber's PPPoE session, and limits the unicast bandwidth
usage, to leave enough bandwidth for multicast traffic. This is useful when the multicast traffic and unicast
traffic share the same physical link to the subscriber in the last mile, when the multicast and unicast traffic
are forwarded onto the last mile link by different devices. This feature is configured on BNG that forwards
the unicast traffic to the subscriber. Based on the IGMP reports received, BNGinforms the unicast QoS shaper
Release 4.3.x
234 OL-28375-03
Multicast Replication
on the PPPoE session to alter the bandwidth limit allowed for unicast traffic flows. Using this HQoS correlation
feature, a service provider can protect the multicast traffic to the PPPoE subscriber from bursty unicast traffic.
The bandwidth profiles for multicast flows need to be configured on BNG.
To define the bandwidth profile, see Configuring Minimum Unicast Bandwidth, on page 235.
To specify the mode for Multicast HQoS, see Configuring Multicast HQOS Correlation Mode or Passive
Mode, on page 236.
Configuring Minimum Unicast Bandwidth
A minimum unicast bandwidth can be configured, to prevent unicast traffic from being completely cut off by
oversubscribed multicast traffic. Perform this task to set the guaranteed minimum unicast bandwidth for a
subscriber using QoS.
SUMMARY STEPS
1. configure
2. dynamic-template
3. type [ppp|ip-subscriber|service]name
4. qos output minimum-bandwidth range
5. exit
DETAILED STEPS
Example:
Step 1
Enters dynamic template configuration mode. dynamic-template
Example:
Step 2
. type [ppp|ip-subscriber|service]name Step 3
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template)# type ppp
p1
Specifies the type of dynamic template that needs
to be applied. Three available types are:
PPP
IP-subscriber
Service
4.3.x
OL-28375-03 235
HQoS Correlation
Sets the guaranteed minimum bandwidth, in kbps,
for a subscriber. Range is from 1 to 4294967295.
qos output minimum-bandwidth range
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template-type)# qos
output minimum-bandwidth 10
Step 4
Exits from the current mode. exit
Example:
RP/0/RSP0/CPU0:router(config-dynamic-template-type)# exit
Step 5
Configuring Minimum Bandwidth: An example
configure
dynamic-template
type ppp p1
service-policy output pmap
multicast ipv4 qos-correlation
qos output minimum-bandwidth 10
end
Configuring Multicast HQOS Correlation Mode or Passive Mode
Perform this task to configure multicast in HQoS correlation mode or passive mode to enable multicast
replication over PPPoE interfaces.
SUMMARY STEPS
1. configure
2. dynamic-template
3. type ppp dynamic-template name
4. multicast ipv4 <qos-correlation | passive>
Release 4.3.x
236 OL-28375-03
HQoS Correlation
DETAILED STEPS
Example:
Step 1
Enter the dynamic-template configuration mode. dynamic-template
Example:
Step 2
Enters the ppp type mode to configure igmp for
type ppp dynamic-template name
Example:
ppp foo
Step 3
Configures the subscriber either in the QoS-correlation
mode (IGMP-HQOS correlation), or passive mode
(multicast forwarding).
multicast ipv4 <qos-correlation | passive>
Example:
Step 4
Configuring Multicast HQOS Correlation Mode: An example
dynamic-template type ppp foo
!
!
end
IGMP to Unicast QoS Shaper Correlation
The Unicast QoS Shaper correlation feature configures the bandwidth profiles for the multicast flows and
allows the IGMP messages to derive the multicast bandwidth usage for each subscriber. On the PPPoE
4.3.x
OL-28375-03 237
subscriber sessions, the amount of multicast bandwidth that a subscriber uses is deducted from the unicast
QoS shaper until a minimum threshold is reached.
For more information about configuring the IGMP QoS shaper, see Configuring the IGMP to HQoS Correlation
Feature in a VRF, on page 238. For more information about configuring the IGMP for subscriber interfaces,
see Configuring IGMP Parameters for Subscriber Interfaces, on page 241.
IGMP uses route-policies to distribute the absolute rate for all multicast flows. For more information for
configuring the route-policy for unicast QoS shaper, see Configuring route-policy for Unicast QoS Shaper,
on page 239.
Configuring the IGMP to HQoS Correlation Feature in a VRF
Perform this task to configure the IGMP to HQoS Correlation Feature in a VRF.
SUMMARY STEPS
1. configure
2. router igmp
3. unicast-qos-adjust adjustment-delay time
4. unicast-qos-adjust download-interval time
5. unicast-qos-adjust holdoff time
6. vrf vrf-name
7. traffic profile profile-name
DETAILED STEPS
Example:
Step 1
Enters the router process for IGMP configuration mode. router igmp
Example:
RP/0/RSP0/CPU0:router(config)# router igmp
Step 2
Configures the time to wait before programming rate in IGMP
QoS shaper for subscriber unicast traffic. The time to wait ranges
from 0 to 10 seconds.
unicast-qos-adjust adjustment-delay time
Example:
RP/0/RSP0/CPU0:router(config-igmp)#
unicast-qos-adjust adjustment-delay 1
Step 3
Configures the time before downloading a batch of interfaces to
IGMP QOS shaper for subscriber unicast traffic. The download
interval time ranges from 10 to 500 milliseconds.
unicast-qos-adjust download-interval time
Example:
unicast-qos-adjust download-interval 10
Step 4
Release 4.3.x
238 OL-28375-03
Configures the hold-off time before QoS clears the stale entries
for the IGMP QoS shaper. The hold-off time ranges from 5 to
1800 seconds.
unicast-qos-adjust holdoff time
Example:
unicast-qos-adjust holdoff 5
Step 5
Enters the VRF configuration mode. vrf vrf-name
Example:
RP/0/RSP0/CPU0:router(config-igmp)# vrf
vrf1
Step 6
Configures the route-policy to be used to map the bandwidth
profile.
traffic profile profile-name
Example:
RP/0/RSP0/CPU0:router(config-igmp-vrf1)#
traffic profile routepolicy1
Step 7
Configuring the IGMP QOS Shaper: An Example
configure
router igmp
unicast-qos-adjust adjustment-delay 1
unicast-qos-adjust download-interval 10
unicast-qos-adjust holdoff 5
vrf vrf1
!
!
end
Configuring route-policy for Unicast QoS Shaper
Perform this task to configure route-policy for unicast QoS shaper.
4.3.x
OL-28375-03 239
SUMMARY STEPS
1. configure
2. router igmp
3. vrf vrf-name
4. traffic profile profile-name
6. show igmp unicast-qos-adjust statistics
7. show igmp unicast-qos-adjust statistics interface interface-name
DETAILED STEPS
Example:
Step 1
Enter the router process for igmp configuration mode. router igmp
Example:
RP/0/RSP0/CPU0:router(config)# router igmp
Step 2
Enters the vrf configuration mode. vrf vrf-name
Example:
RP/0/RSP0/CPU0:router(config-igmp)# vrf
vrf1
Step 3
Configures the route-policy to be used to map the bandwidth
profile.
traffic profile profile-name
Example:
Step 4
Release 4.3.x
240 OL-28375-03
(Optional) Displays the features's internal statistics, such as total
number of interface groups under adjustment, uptime since last
show igmp unicast-qos-adjust statistics
Example:
RP/0/RSP0/CPU0:router# show igmp
unicast-qos-adjusted statistics
Step 6
clear command, and total number of rate adjustment calls for
unicast QoS shaper.
(Optional) Displays the interface name, number of flows adjusted,
total rate adjusted, uptime after first adjustment for unicast QoS
shaper.
showigmp unicast-qos-adjust statistics interface
interface-name
Example:
unicast-qos-adjusted statistics interface
interface1
Step 7
Configuring route-policy for Unicast QoS Shaper: Examples
#Adding a route-policy for profile1
route-policy profile1
if destination in (239.0.0.0/8 le 32) then
set weight 1000
endif
end-policy
# Configuring profile1 for Unicast QoS Shaper
router igmp
vrf vrf1
traffic profile profile1
!
!
end
Configuring IGMP Parameters for Subscriber Interfaces
Perform this task to configure IGMP parameters for subscriber interfaces.
SUMMARY STEPS
1. configure
2. dynamic-template
3. type ppp dynamic-template name
4. igmp explicit-tracking
5. igmp query-interval value
6. igmp query-max-response-time query-response-value
4.3.x
OL-28375-03 241
DETAILED STEPS
Example:
Step 1
Enter the dynamic-template configuration mode. dynamic-template
Example:
Step 2
Enters the ppp type mode to configure igmp for
type ppp dynamic-template name
Example:
ppp foo
Step 3
Enables IGMPv3 explicit host tracking. igmp explicit-tracking
Example:
igmp explicit-tracking
Step 4
Sets the query-interval in seconds for igmp. igmp query-interval value Step 5
Example:
igmp query-interval 60
The igmp query-interval value, in seconds,
should be in the range from 1 to 3600. With
16000 PPPoE subscribers or less, the
recommended value, that also the default, is 60
seconds.
Note
Sets the query-max-response-time in seconds for igmp. igmp query-max-response-time query-response-value Step 6
Example:
igmp query-max-response-time 4
The igmp query-interval value, in seconds, is in
the range from 1 to 12.
Note
Release 4.3.x
242 OL-28375-03
Configuring IGMP for Subscriber Interfaces: An example
dynamic-template type ppp foo
igmp explicit-tracking
igmp query-interval 60
igmp query-max-response-time 4
!
!
end
IGMP Accounting
The Internet Group Management Protocol (IGMP) accounting feature enables BNG to maintain a statistics
file to log the instances of subscriber joining, or leaving a multicast group. The file's format is:
harddisk:/usr/data/igmp/accounting.dat.<Node ID>.<YYMMDD>
where
Node ID is the name of the node that generates the file; for example, RP/0/RSP0/CPU0.
YY is the year, MM is the month, and DD is the day.
An example of the statistics file name is:
harddisk:/usr/data/igmp/accounting.dat.RP_0_RSP0_CPU0.101225
The statistics file is stored on the route processor (RP) that is active. If a failover event occurs, then a new file
is created on the new active RP, and no attempt is made to mirror the data between the active and the standby
RP. Thus, the statistics files must be retrieved from both the active and standby RPs.
By default, the IGMP Accounting feature adds one file each day. To avoid exhausting disk space, you can
specify in howmany files, or for howmany days, data should be retained, see Configuring IGMP Accounting,
on page 243. Files older than the specified period are deleted, and the data is discarded from BNG. The
maximum size of each file should be no more than 250 MB.
Configuring IGMP Accounting
Perform this task to configure the IGMP accounting.
SUMMARY STEPS
1. configure
2. router igmp
3. accounting [ max-history ] days
5. show igmp interface
4.3.x
OL-28375-03 243
IGMP Accounting
DETAILED STEPS
Example:
Step 1
Enter the router process for IGMP configuration mode. router igmp
Example:
RP/0/RSP0/CPU0:router(config)# router
igmp
Step 2
Configures the IGMP accounting. The max-history parameter is
optional and specifies how many files are kept; this number is
equivalent to the number of days in the history.
accounting [ max-history ] days
Example:
accounting max-history 50
Step 3
session.
(Optional) Displays IGMP interface information. show igmp interface
Example:
interface
Step 5
Configuring IGMP Accounting: An example
configure
router igmp
!
!
end
Release 4.3.x
244 OL-28375-03
IGMP Accounting
DAPS Support
A Distributed Address Pool Service (DAPS) allows address pools to be shared between DHCP processes that
run on a line card (LC) and the route processor (RP). The DHCP Server and PPPoE subscribers are clients to
DAPS, and are known as the DAPS client. DAPS is used to return IP address to clients only when the RADIUS
attributes contain the attribute "Pool Name". If the RADIUS attribute for a subscriber contains a fixed address,
then the client does not contact DAPS for its IP address.
DAPS runs in two forms, as DAPS server on the RP, and as DAPS-Proxy on the LC. The RP has an in-build
DAPS-Proxy module. This model ensures that all DAPS clients always talk to the DAPS-Proxy. The
DAPS-Proxy instances talk to the central DAPS-Server on the RP for address assignments and other requests.
DAPS-Proxy runs on all the LCs in the system. The DAPS-Proxy running on an LC can service multiple
clients, from that LC; for example, PPP, DHCPv6, IPv6ND. DAPS serves multiple DAPS clients on two or
more nodes. A separate DAPS-Proxy process runs on each node and connects locally to each DAPS Client.
DAPS supports dynamic IPv4 and IPv6 address allocation by pool name. For more information about
configuring IPv4 DAPS, see Configuring IPv4 Distributed Address Pool Service, on page 245. To create a
configuration pool for IPv6, see Creating a Configuration Pool Submode, on page 246.
You can configure various DAPS IPv6 parameters in the IPv6 configuration submode. You can configure the
subnet number and mask for an IPv6 address pool, for more information, see Configuring the Subnet Number
and Mask for an Address Pool, on page 248. You can specify parameters such as a range of IPv6 addresses.
For more information, see Specifying a Range of IPv6 Addresses, on page 250. To specify a utilization
threshold, see Specifying a Utilization Threshold, on page 251. To specify a set of prefixes or addresses inside
a subnet, see Specifying a Set of Addresses or Prefixes Inside a Subnet, on page 254. You can also specify
the length of a prefix. For more information, see Specifying the Length of the Prefix, on page 253.
Configuring IPv4 Distributed Address Pool Service
Perform this task to configure IPv4 distributed address pool service (DAPS).
SUMMARY STEPS
1. configure
2. pool ipv4 ipv4-pool-name
3. address-range first_address second_address
4. pool vrf vrf-name ipv4 ipv4-pool-name{address-range address-range}
DETAILED STEPS
Example:
Step 1
4.3.x
OL-28375-03 245
DAPS Support
Configures IPv4 pool name. pool ipv4 ipv4-pool-name
Example:
RP/0/RSP0/CPU0:router(config)# pool ipv4 pool1
Step 2
Configures the address range for allocation. address-range first_address second_address
Example:
RP/0/RSP0/CPU0:router(config-pool-ipv4)# address-range
1.1.1.1 9.8.9.8
Step 3
Configures IPv4 pool name. pool vrf vrf-name ipv4 ipv4-pool-name{address-range
address-range}
Step 4
Example:
RP/0/RSP0/CPU0:router(config)# pool vrf vrf1 ipv4 pool1
address-range 1.1.1.1 9.8.9.8
Configuring IPv4 Distributed Address Pool Service: An example
pool ipv4 pool1
address-range 1.1.1.1 9.8.9.8
pool vrf vrf1 ipv4 pool1 address-range 1.1.1.1 9.8.9.8
!
!
end
Creating a Configuration Pool Submode
Performthis task to create and enable an IPv6 configuration pool submode for a default VRF and for a specific
VRF.
Release 4.3.x
246 OL-28375-03
SUMMARY STEPS
1. configure
2. pool ipv6 ipv6-pool-name
4. configure
5. pool vrf vrf_name ipv6 ipv6-pool-name
DETAILED STEPS
Example:
Step 1
Creates the IPv6 pool name for a default VRF and enters the pool IPv6
configuration submode.
pool ipv6 ipv6-pool-name
Example:
RP/0/RSP0/CPU0:router(config)# pool
ipv6 pool1
Step 2
session.
Example:
Step 4
Creates the IPv6 pool name for a specific VRF and enters the pool IPv6
configuration submode.
pool vrf vrf_name ipv6 ipv6-pool-name
Example:
RP/0/RSP0/CPU0:router(config)# pool
vrf vrf1 ipv6 pool1
Step 5
4.3.x
OL-28375-03 247
session.
Creating a Configuration Pool Submode: An example
configure
pool ipv6 pool1 (default vrf)
!
!
configure
pool vrf vrf1 ipv6 pool1 (for a specific vrf)
!
!
end
Configuring the Subnet Number and Mask for an Address Pool
Perform this task to create the subnet number and mask for an IPv6 address pool.
SUMMARY STEPS
1. configure
3. prefix-length value
4. network subnet
5. utilization-mark high value low value
6. exclude low_ip_address high_ip_address
Release 4.3.x
248 OL-28375-03
DETAILED STEPS
Example:
Step 1
Creates the IPv6 pool name for a specific VRF and
enters the pool IPv6 configuration submode.
Example:
RP/0/RSP0/CPU0:router(config)# pool vrf default ipv6 test
Step 2
Specifies the length of the prefix that is assigned to
the clients. The value of the prefix length ranges from
1 to 128.
prefix-length value
Example:
RP/0/RSP0/CPU0:router(config-pool-ipv6)# prefix-length
120
Step 3
Specifies a set of addresses or prefixes inside a
subnet.
network subnet
Example:
RP/0/RSP0/CPU0:router(config-pool-ipv6)# network
1101:1::/114
Step 4
The prefix-length command must be
mandatorily configured whenever the
network command is used.
Note
Specifies a utilization threshold in the pool IPv6
submode. The high and low values are represented
as percentages between 0 and 100.
utilization-mark high value low value
Example:
RP/0/RSP0/CPU0:router(config-pool-ipv6)# utilization-mark
high 70 low 30
Step 5
Specifies a range of IPv6 addresses or prefixes that
DAPS must not assign to clients. The high and low
exclude low_ip_address high_ip_address
Example:
RP/0/RSP0/CPU0:router(config-pool-ipv6)# exclude
1101:1::100 ::
Step 6
values are represented as percentages between 0 and
100.
Multiple exclude commands are allowed
within a pool. To exclude a single address,
<high_ip_address> can be omitted.
Note
4.3.x
OL-28375-03 249
Configuring the Subnet Number and Mask for an Address Pool: An example
configure
pool vrf default ipv6 test
prefix-length 120
network 1101:1::/114
utilization-mark high 70 low 30
exclude 1101:1::100 ::
!
!
end
Specifying a Range of IPv6 Addresses
Perform this task to specify a range of IPv6 addresses within a pool.
SUMMARY STEPS
1. configure
3. address-range low_ip_address high_ip_address
DETAILED STEPS
Example:
Step 1
Example:
RP/0/RSP0/CPU0:router(config)# pool vrf vrf1 ipv6
addr_vrf
Step 2
Release 4.3.x
250 OL-28375-03
Specifying a Range of IPv6 Addresses
Specifies the range of IPv6 addresses within a pool.
Multiple address-ranges are allowed within a pool.
address-range low_ip_address high_ip_address
Example:
RP/0/RSP0/CPU0:router(config-pool-ipv6)# address-range
1234::2 1234::3e81
Step 3
Specifying a Range of IPv6 Addresses: An example
configure
pool vrf vrf1 ipv6 addr_vrf
address-range 1234::2 1234::3e81
!
!
end
Specifying a Utilization Threshold
Perform this task to specify a utilization threshold for a specific VRF in the pool IPv6 submode.
SUMMARY STEPS
1. configure
4. network subnet
4.3.x
OL-28375-03 251
DETAILED STEPS
Example:
Step 1
Example:
Step 2
the clients. The value of the prefix length ranges
from 1 to 128.
prefix-length value
Example:
120
Step 3
subnet.
network subnet
Example:
1101:1::/114
Step 4
The prefix-length command should be
Note
Example:
high 70 low 30
Step 5
Specifying a Utilization Threshold: An example
configure
Release 4.3.x
252 OL-28375-03
prefix-length 120
network 1101:1::/114
!
!
end
Specifying the Length of the Prefix
Perform this task to specify the length of the prefix that is assigned to the clients.
SUMMARY STEPS
1. configure
4. prefix-range low_ipv6_prefix high_ipv6_prefix
DETAILED STEPS
Example:
Step 1
Example:
RP/0/RSP0/CPU0:router(config)# pool vrf vrf1 ipv6
prefix_vrf
Step 2
Specifies the length of the prefix that is assigned to the
clients. The value of the prefix length ranges from 1 to
128.
prefix-length value
Example:
64
Step 3
Specifies a range of IPv6 address prefixes for a specific
VRF in the pool IPv6 configuration mode.
prefix-range low_ipv6_prefix high_ipv6_prefix
Example:
RP/0/RSP0/CPU0:router(config-pool-ipv6)# prefix-range
9fff:1:: 9fff:1:0:3e7f::
Step 4
The prefix-length must be mandatorily
configured whenever prefix-range is
configured.
Note
4.3.x
OL-28375-03 253
Specifying the Length of the Prefix
Specifying the Length of the Prefix that is Assigned to the Clients: An example
configure
pool vrf vrf1 ipv6 prefix_vrf
prefix-length 64
prefix-range 9fff:1:: 9fff:1:0:3e7f::
!
!
end
Specifying a Set of Addresses or Prefixes Inside a Subnet
Perform this task to specify a set of addresses or prefixes inside a subnet in the pool IPv6 configuration
submode.
SUMMARY STEPS
1. configure
4. network subnet
6. exclude low_ip_address high_ip_address
DETAILED STEPS
Example:
Step 1
Release 4.3.x
254 OL-28375-03
Example:
Step 2
the clients. The value of the prefix length ranges from
1 to 128.
prefix-length value
Example:
120
Step 3
subnet.
network subnet
Example:
1101:1::/114
Step 4
The prefix-length command should be
Note
Example:
high 70 low 30
Step 5
Specifies a range of IPv6 addresses or prefixes that
DAPS must not assign to clients. The high and low
exclude low_ip_address high_ip_address
Example:
RP/0/RSP0/CPU0:router(config-pool-ipv6)# exclude
1101:1::100 ::
Step 6
values are represented as percentages between 0 and
100.
Multiple exclude commands are allowed
within a pool. To exclude a single address,
<high_ip_address> can be omitted.
Note
4.3.x
OL-28375-03 255
Specifying a Set of Addresses or Prefixes Inside a Subnet: An example
configure
prefix-length 120
network 1101:1::/114
exclude 1101:1::100 ::
!
!
end
HTTP Redirect Using PBR
The HTTP Redirect (HTTPR) feature is used to redirect subscriber traffic to a destination other than the one
to which it was originally destined. The HTTPR feature is implemented using Policy Based Routing (PBR)
that makes packet forwarding decisions based on the policy configuration, instead of routing protocols. The
HTTPRfeature is implemented by sending an HTTP redirect response, which contains the redirect URL, back
to the HTTP client that originally sent the request. Thereafter, the HTTP client sends requests to the redirected
URL. HTTPR is supported for both IPv4 and IPv6 subscribers.
The most common use of HTTPR feature is for initial logon. In some cases, it is not possible to uniquely
identify a subscriber and authorize them. This happens when the subscriber is using a shared network access
mediumto connect to the network. In such cases, the subscriber is allowed to access the network but restricted
to what is known as an "open-garden". An open-garden is a collection of network resources that subscribers
can access as long as they have physical access to the network. Subscribers do not have to provide authentication
information before accessing the web sites in an open-garden.
When subscribers try to access resources outside the open-garden (which is called the "walled-garden"), they
are redirected to a web logon portal. The walled-garden refers to a collection of web sites or networks that
subscribers can access after providing minimal authentication information. The web logon portal requires the
subscriber to login using a username and password. Thereafter, the web logon portal sends an account-logon
CoAto BNGwith user credentials. On successful authentication of these credentials, BNGdisables the redirect
and applies the correct subscriber policies for direct network access. Other uses of HTTPR include periodic
redirection to a web portal for advertising reasons, redirection to a billing server, and so on.
The PBRfunction is configured in its own dynamic template. If the dynamic template contains other functions
too, then the PBR policy that redirects packets must be deactivated using a CoA.
BNGmaintains HTTP redirect statistics counters that track the number of packets that are being either redirected
or dropped. The HTTP protocol uses some status codes to implement HTTPR. Currently, the redirect codes
302 (for HTTP version 1.0) and 307 (for HTTP version 1.1) are supported on BNG.
Note HTTP redirect applies only to HTTP packets. As a result, other services such as SMTP, FTP are not
affected by this feature. Nevertheless, if these other services are part of the redirect classification
rules, then the packets are dropped and not forwarded.
HTTPS is not supported.
Destination URL-based classification is not supported.
The process of configuring HTTPR involves these stages:
Release 4.3.x
256 OL-28375-03
HTTP Redirect Using PBR
Creating access lists that define the redirected and open-garden permissions. See, Identifying HTTP
Destinations for Redirection, on page 257.
Creating the class-maps that uses the access list to classify the traffic as redirected, or permitted to access
open-garden. See, Configuring Class Maps for HTTP Redirection, on page 261.
Creating the policy-map to define the action to be performed on the traffic classified using class-maps.
See, Configuring Policy Map for HTTP Redirect, on page 263.
Creating the dynamic template to apply the service policy. See Configuring Dynamic Template for
Applying HTTPR Policy, on page 266.
To configure a web logon that specifies a time limit to perform the authentication, see Configuring Web
Logon, on page 267.
Identifying HTTP Destinations for Redirection
Perform this task to define access lists that identify http destinations that require redirection or are part of an
open garden:
4.3.x
OL-28375-03 257
SUMMARY STEPS
1. configure
2. {ipv4 | ipv6}access-list redirect_acl_name
3. Do one of the following:
[ sequence-number]{permit | deny} source source-wildcard destination destination-wildcard
[precedence precedence] [dscp dscp] [fragments] [packet-length operator packet-length value]
[log | log-input]
[ sequence-number ] {permit | deny} protocol {source-ipv6-prefix/prefix-length | any | host
source-ipv6-address} [operator {port | protocol-port}] {destination-ipv6-prefix/prefix-length | any
| host destination-ipv6-address} [operator {port | protocol-port}] [dscp value] [routing] [authen]
[destopts] [fragments] [packet-length operator packet-length value] [log | log-input]
4. Repeat Step 3 as necessary, adding statements by sequence number. Use the no sequence-number
command to delete an entry.
5. {ipv4 | ipv6}access-list open_garden_acl
6. Do one of the following:
[ sequence-number]{permit | deny} source source-wildcard destination destination-wildcard
[precedence precedence] [dscp dscp] [fragments] [packet-length operator packet-length value]
[log | log-input]
[ sequence-number ] {permit | deny} protocol {source-ipv6-prefix/prefix-length | any | host
source-ipv6-address} [operator {port | protocol-port}] {destination-ipv6-prefix/prefix-length | any
| host destination-ipv6-address} [operator {port | protocol-port}] [dscp value] [routing] [authen]
[destopts] [fragments] [packet-length operator packet-length value] [log | log-input]
7. Repeat Step 6 as necessary, adding statements by sequence number. Use the no sequence-number
command to delete an entry.
DETAILED STEPS
Example:
Step 1
Enters either IPv4 or IPv6 access list configuration mode and
configures the named access list.
{ipv4 | ipv6}access-list redirect_acl_name
Example:
RP/0/RSP0/CPU0:router(config)# ipv4 access-lists
redirect_acl
Step 2
Release 4.3.x
258 OL-28375-03
or
redirect_acl
Specifies one or more conditions allowed or denied in IPv4
or IPv6 access list redirect_acl.
Do one of the following: Step 3
[ sequence-number]{permit | deny} source
source-wildcard destination destination-wildcard The optional log keyword causes an information
logging message about the packet that matches the entry
to be sent to the console.
[precedence precedence] [dscp dscp] [fragments]
[packet-length operator packet-length value] [log |
log-input]
The optional log-input keyword provides the same
function as the log keyword, except that the logging
message also includes the input interface.
[ sequence-number ] {permit | deny} protocol
{source-ipv6-prefix/prefix-length | any | host
source-ipv6-address} [operator {port | protocol-port}]
or {destination-ipv6-prefix/prefix-length | any | host
destination-ipv6-address} [operator {port |
access list redirect_acl.
protocol-port}] [dscp value] [routing] [authen]
[destopts] [fragments] [packet-length operator
packet-length value] [log | log-input] Refer to the deny (IPv6) and permit (IPv6) commands
for more information on filtering IPv6 traffic based on
based on IPv6 option headers and optional, upper-layer
protocol type information.
Example:
RP/0/RSP0/CPU0:router(config-ipv4-acl)# 10 permit
Every IPv6 access list has an implicit deny ipv6 any
any statement as its last match condition. An IPv6
access list must contain at least one entry for the
implicit deny ipv6 any any statement to take effect.
Note
172.16.0.0 0.0.255.255
RP/0/RSP0/CPU0:router(config-ipv4-acl)# 20 deny
192.168.34.0 0.0.0.255
or
icmp any any
RP/0/RSP0/CPU0:router(config-ipv6-acl)# 30 deny tcp
any any gt 5000
Allows you to revise an access list. Repeat Step 3 as necessary, adding statements by sequence
number. Use the no sequence-number command to delete
an entry.
Step 4
Enters either IPv4 or IPv6 access list configuration mode and
configures the named access list for open garden.
{ipv4 | ipv6}access-list open_garden_acl
Example:
open_garden_acl
Step 5
or
open_garden_acl
access list open_garden_acl.
Do one of the following: Step 6
[ sequence-number]{permit | deny} source
source-wildcard destination destination-wildcard
4.3.x
OL-28375-03 259
The optional log keyword causes an information
logging message about the packet that matches the entry
to be sent to the console.
[precedence precedence] [dscp dscp] [fragments]
[packet-length operator packet-length value] [log |
log-input]
The optional log-input keyword provides the same
function as the log keyword, except that the logging
message also includes the input interface.
[ sequence-number ] {permit | deny} protocol
{source-ipv6-prefix/prefix-length | any | host
source-ipv6-address} [operator {port | protocol-port}]
{destination-ipv6-prefix/prefix-length | any | host
or
destination-ipv6-address} [operator {port |
protocol-port}] [dscp value] [routing] [authen]
access list open_garden_acl.
[destopts] [fragments] [packet-length operator
packet-length value] [log | log-input]
Refer to the deny (IPv6) and permit (IPv6) commands
for more information on filtering IPv6 traffic based on
Example:
based on IPv6 option headers and optional, upper-layer
protocol type information.
172.16.0.0 0.0.255.255
Every IPv6 access list has an implicit deny ipv6 any
any statement as its last match condition. An IPv6
access list must contain at least one entry for the
implicit deny ipv6 any any statement to take effect.
Note
RP/0/RSP0/CPU0:router(config-ipv4-acl)# 20 deny
192.168.34.0 0.0.0.255
or
icmp any any
RP/0/RSP0/CPU0:router(config-ipv6-acl)# 30 deny tcp
any any gt 5000
Allows you to revise an access list. Repeat Step 6 as necessary, adding statements by sequence
number. Use the no sequence-number command to delete
an entry.
Step 7
Identifying HTTP Destinations for Redirection: An example
configure
ipv4 access-list <redirect-acl>
10 permit tcp any any syn eq www
20 permit tcp any any ack eq www
30 permit tcp any any eq www
ipv4 access-group <allow-acl>
Release 4.3.x
260 OL-28375-03
10 permit tcp any 10.1.1.0 0.0.0.255 eq www
40 permit udp any any eq domain
!
!
!
end
configure
ipv6 access-list <redirect-acl>
10 permit tcp any any syn eq www
20 permit tcp any any ack eq www
30 permit tcp any any eq www
ipv6 access-group <allow-acl>
40 permit udp any any eq domain
!
!
!
end
Configuring Class Maps for HTTP Redirection
Perform this task to configure the class maps for HTTP redirection. It makes use of previously defined ACLs.
Before You Begin
The configuration steps mentioned in Identifying HTTP Destinations for Redirection, on page 257 has to be
completed before performing the configuration of the HTTPR class maps.
SUMMARY STEPS
1. configure
2. class-map type traffic match-all open-garden-class_name
3. match [not] access-group{ipv4 | ipv6} open_garden_acl
4. end-class-map
5. class-map type traffic match-all http_redirect-class_name
6. match [not] access-group {ipv4 | ipv6} redirect_acl
7. end-class-map
DETAILED STEPS
Example:
Step 1
4.3.x
OL-28375-03 261
Defines a traffic class and the associated rules that match
packets to the class for a open garden class.
class-map type traffic match-all open-garden-class_name
Example:
traffic match-all CL1
Step 2
Identifies a specified access control list (ACL) number as
the match criteria for a class map.
match [not] access-group{ipv4 | ipv6} open_garden_acl
Example:
RP/0/RSP0/CPU0:router(config-cmap)# match not
access-group ipv4 open_garden_acl
Step 3
The redirect acl name provided in this step is the
one configured in the configuration step mentioned
in the prerequisites.
Note
or
access-group ipv6 open_garden_acl
Ends the configuration of match criteria for the class and
exits the class map configuration mode.
end-class-map
Example:
Step 4
Defines a traffic class and the associated rules that match
packets to the class for a open garden class.
class-map type traffic match-all http_redirect-class_name
Example:
traffic match-all RCL1
Step 5
Identifies a specified access control list (ACL) number as
the match criteria for a class map.
match [not] access-group {ipv4 | ipv6} redirect_acl
Example:
access-group ipv4 redirect-acl
Step 6
The redirect acl name provided in this step is the
one configured in the configuration step mentioned
in the prerequisites.
Note
or
access-group ipv6 redirect-acl
Ends the configuration of match criteria for the class and
exits the class map configuration mode.
end-class-map
Example:
Step 7
Release 4.3.x
262 OL-28375-03
Configuring Class Maps for HTTP Redirection: An example
configure
class-map type traffic [match-any | match-all] <open-garden-class>
match [not] access-group ipv4 allow-acl
end-class-map
class-map type traffic [match-any | match-all] <http-redirect-class>
match [not] access-group ipv4 redirect-acl
end-class-map
!
!
!
end
configure
class-map type traffic [match-any | match-all] <open-garden-class>
match [not] access-group ipv6 allow-acl
end-class-map
class-map type traffic [match-any | match-all] <http-redirect-class>
match [not] access-group ipv6 redirect-acl
end-class-map
!
!
!
end
Configuring Policy Map for HTTP Redirect
Perform this task to configure policy maps for http redirect.
Before You Begin
The configuration steps mentioned in Identifying HTTP Destinations for Redirection, on page 257 and
Configuring Class Maps for HTTP Redirection, on page 261 have to be completed before performing the
configuration of the policy-map for HTTPR.
4.3.x
OL-28375-03 263
SUMMARY STEPS
1. configure
2. policy-map type pbr http-redirect_policy_name
3. class type traffic open_garden_class_name
4. transmit
5. class type traffic http_redirect-class_name
6. http-redirect redirect_url
8. drop
9. end-policy-map
DETAILED STEPS
Example:
Step 1
Creates a policy map of type policy-based routing that
can be attached to one or more interfaces to specify a
service policy.
policy-map type pbr http-redirect_policy_name
Example:
RP/0/RSP0/CPU0:router(config)# policy-map type pbr
RPL1
Step 2
Specifies the name of the class whose policy you want to
create or change.
class type traffic open_garden_class_name
Example:
traffic CL1
Step 3
The open garden acl name provided in this step
is the one configured in the configuration step
mentioned in the prerequisites.
Note
Forwards the packet to the original destination. transmit
Example:
RP/0/RSP0/CPU0:router(config-pmap)# transmit
Step 4
Specifies the name of the class whose policy you want to
create or change.
class type traffic http_redirect-class_name
Example:
traffic RCL1
Step 5
The open garden acl name provided in this step
Note
Release 4.3.x
264 OL-28375-03
Specifies the URL to which the HTTP requests should
be redirected.
http-redirect redirect_url
Example:
RP/0/RSP0/CPU0:router(config-pmap-c)# http-redirect
redirect_url
Step 6
Configures default classes that cannot be used with
user-defined classes.
class class-default
Example:
class-default
Step 7
Drops the packet. drop
Example:
RP/0/RSP0/CPU0:router(config-pmap)# drop
Step 8
Ends the configuration of a policy map and exits the
policy map configuration mode.
end-policy-map
Example:
RP/0/RSP0/CPU0:router(config-cmap)# end-policy-map
Step 9
Configuring Policy Map for HTTP Redirect: An example
configure
policy-map type pbr <http-redirect-policy>
class type traffic <open-garden-class>
transmit
!
class type traffic <http-redirect-class>
http-redirect <redirect-url>
!
class class-default
drop
!
end-policy-map
!
!
4.3.x
OL-28375-03 265
!
end
Configuring Dynamic Template for Applying HTTPR Policy
Perform this task to configure dynamic template for applying the HTTPR policy to subscriber sessions.
Before You Begin
The configuration steps mentioned in Configuring Policy Map for HTTP Redirect, on page 263 have to be
completed before defining the dynamic template that uses a previously defined policy-map.
Ensure that the Dynamic template contains only the Policy Based Routing policy, so it can be easily
deactivated after web login.
Note
SUMMARY STEPS
1. configure
2. dynamic-template type ipsubscriber redirect_template_name
3. service-policy type pbr http-redirect-policy
DETAILED STEPS
Example:
Step 1
Creates a dynamic template of type "ipsubscriber". dynamic-template type ipsubscriber
redirect_template_name
Step 2
Example:
type ipsubscriber RDL1
Attaches the service policy as a pbr type within a policy map
created in the earlier configuration.
service-policy type pbr http-redirect-policy
Example:
RP/0/RSP0/CPU0:router(config-pmap)# service-policy
type pbr RPL1
Step 3
The http redirect policy name provided in this step
Note
Release 4.3.x
266 OL-28375-03
Configuring Dynamic Template for Applying HTTPR Policy
Configuring Dynamic Template for Applying HTTPR Policy: An example
configure
dynamic-template type ip <redirect-template>
service-policy type pbr <http-redirect-policy>
!
!
!
end
Configuring Web Logon
Perform this task to configure Web Logon. As an example, a timer defines the maximum time permitted for
authentication.
4.3.x
OL-28375-03 267
SUMMARY STEPS
1. configure
2. class-map type control subscriber match-all classmap_name
3. match timer name
4. match authen-status authenticated
5. policy-map type control subscriber policymap_name
8. sequence_number activate dynamic-template dt_name
9. sequence_number activate dynamic-template dt_name
10. sequence_number set-timer timer_name value
11. event account-logon match-all
14. sequence_number deactivate dynamic-templatedt_name
15. sequence_number stop-timer timer_name
16. event time-expiry match-all
17. class type control subscriber class_name do-all
18. sequence_number disconnect
DETAILED STEPS
Example:
Step 1
Configures a subscriber control class-map with the
match-all match criteria.
class-map type control subscriber match-all
classmap_name
Example:
control subscriber match-all IP_UNATH_COND
Step 2
Configures a match criteria for the class along with timer
details.
match timer name
Example:
RP/0/RSP0/CPU0:router(config-cmap)# match timer
AUTH_TIMER
Step 3
Release 4.3.x
268 OL-28375-03
Configures a match criteria for the class along with
authentication status details.
match authen-status authenticated
Example:
RP/0/RSP0/CPU0:router(config-cmap)# match timer
AUTH_TIMER
Step 4
Configures a subscriber control policy-map. policy-map type control subscriber policymap_name
Example:
control subscriber match-all RULE_IP_WEBSESSION
Step 5
Configures the session start policy event that runs all
the matched classes.
Example:
Step 6
Configures the class to which the subscriber is to be
matched. When there is a match, execute all actions that
follow until a failure is encountered.
Example:
control subscriber class-default do-until-failure
Step 7
Activates the dynamic-template defined locally on the
CLI with the specified dynamic template name.
sequence_number activate dynamic-template dt_name
Example:
dynamic-template DEFUALT_IP_SERVICE
Step 8
Activates the dynamic-template defined locally on the
CLI with the specified dynamic template name.
sequence_number activate dynamic-template dt_name
Example:
dynamic-template HTTP_REDIRECT
Step 9
Sets a timer to run a rule on its expiry. The timer value,
specified in minutes, ranges from 0 to 4294967295.
sequence_number set-timer timer_name value
Example:
RP/0/RSP0/CPU0:router(config-pmap-c)# 10 set-timer
AUTH_TIMER 4567
Step 10
Configures the account logon policy event that runs all
matched classes.
event account-logon match-all
Example:
Step 11
Configures the class to which the subscriber is to be
matched. When there is a match, execute all actions that
follow, until a failure is encountered.
Example:
control subscriber class-default do-until-failure
Step 12
4.3.x
OL-28375-03 269
Specifies and authenticates the default AAA method
list.
Example:
Step 13
Disables the timer before it expires. sequence_number deactivate dynamic-templatedt_name
Example:
RP/0/RSP0/CPU0:router(config-pmap-c)# 10 deactivate
dynamic-template HTTP_REDIRECT
Step 14
Disables the timer before it expires. sequence_number stop-timer timer_name
Example:
RP/0/RSP0/CPU0:router(config-pmap-c)# 20 stop-timer
AUTH_TIMER
Step 15
Configures the timer expiry policy event that runs all
the matched classes.
event time-expiry match-all
Example:
Step 16
Configures the class to which the subscriber has to be
matched. When there is a match, execute all actions.
class type control subscriber class_name do-all
Example:
control subscriber IP_UNAUTH_COND do-all
Step 17
Disconnects the session. sequence_number disconnect
Example:
RP/0/RSP0/CPU0:router(config-pmap-c)# 10 disconnect
Step 18
Release 4.3.x
270 OL-28375-03
Configuring Web Logon: An example
This example illustrates an IP session that is HTTP-redirected to an authentication web-portal for credentials.
On successful authentication, the timer is unset. Otherwise, the subscriber gets disconnected when the timer
window expires:
class-map type control subscriber match-all IP_UNAUTH_COND
match timer AUTH_TIMER
match authen-status unauthenticated
policy-map type control subscriber RULE_IP_WEBSESSION
10 activate dynamic-template DEFAULT_IP_SERVICE
20 activate dynamic-template HTTP_REDIRECT
30 set-timer AUTH_TIMER 5
event account-logon match-all
10 authenticate aaa list default
15 deactivate dynamic-template HTTP_REDIRECT
20 stop-timer AUTH_TIMER
event timer-expiry match-all
class type control subscriber IP_UNAUTH_COND do-all
10 disconnect
These sections provide references related to implementing BNG subscriber features.
MIBs
MIBs Link MIB
4.3.x
OL-28375-03 271
Link Description
Release 4.3.x
272 OL-28375-03
AP P E NDI X A
XML Support for BNG Features
Most BNG features, such as AAA, DHCP, Policy Plane, PPPoE, DAPS, and Subscriber Database support
XML based router configuration. The Cisco XML API can be used to configure routers or request information
about configuration, management, and operation of the routers. For details about using the Cisco XML API,
see the latest release of Cisco IOS XR XML API Guide listed at http://www.cisco.com/en/US/products/ps9853/
products_programming_reference_guides_list.html.
The Cisco XML API uses XML commands to configure the router. The following sections list the supported
XML commands for the BNG features.
AAA XML Support, page 273
DHCP XML Support, page 276
Control Policy XML Support, page 279
DAPS XML Support, page 282
PPPoE XML Support, page 283
Subscriber Database XML Support, page 285
AAA XML Support
The support for XML is available for RADIUS that retrieves the accounting and authorization request statistics.
The mapping between CLI and XML entries for the AAA commands are as follows:
XML CLI
AAA.RADIUS. DeadCriteria.Time radius-server
dead-criteria time
AAA.RADIUS. DeadCriteria.Tries radius-server
dead-criteria tries
AAA.RADIUS. IPv4.DSCP radius-server ipv4
dscp <value>
AAA.RADIUS.Key radius-server key
{0 | 7 | LINE}
4.3.x
OL-28375-03 273
XML CLI
AAA.RADIUS.Retransmit radius-server
retransmit <limit>
AAA.RADIUS.Timeout radius-server
timeout <number>
AAA.RADIUS.SourcePort.Extended radius-server
source-port
extended
AAA.RADIUS.DeadTime radius-server
deadtime
AAA.RADIUS.LoadBalance.Method.LeastOutstanding radius-server
load-balance
method
least-outstanding
AAA.RADIUS.AttributeListTable.AttributeList.Enable radius-server
attribute list
<attribute-name>
AAA.RADIUS.AttributeListTable.AttributeList.Attribute radius-server
attribute list
<attribute-name>
attribute
<radius-attributes>
AAA.RADIUS.VSA.Attribute.Ignore.Unknown radius-server vsa
attribute ignore
unknown
AAA.RADIUS.HostTable.Host.Retransmit Radius-server host
<> retransmit
AAA.RADIUS.HostTable.Host.Timeout Radius-server host
<> timeout
AAA.RADIUS.HostTable.Host.Key radius-server host
<> key {0 | 7 |
LINE}
AAA.RADIUS.DynamicAuthorization.ClientTable.Client.ServerKey aaa server radius
dynamic-author
client
<ip-address> vrf
<vrf-name>
server-key {0 | 7 |
LINE}
AAA.RADIUS.DynamicAuthorization.Ignore aaa server radius
dynamic-author
ignore {server key
| session key }
Release 4.3.x
274 OL-28375-03
AAA XML Support
XML CLI
AAA.RADIUS.DynamicAuthorization.Port aaa server radius
dynamic-author
port <port num>
AAA.AccountingTable.Accounting aaa accounting
system default
start-stop
[broadcast] {group
{radius |
NAME1}} [group
NAME2..] aaa
accounting system
rp-failover default
start-stop
[broadcast] {group
{radius |
NAME1}} [group
NAME2..
AAA.RADIUSAttribute.NASPortID.Format aaa radius
attribute
nas-port-id format
FORMAT_NAME
AAA.ServerGroups.RADIUSServerGroupTable.RADIUSServerGroup.Authorization.Reply aaa group server
radius
<group-name> {
authorization } {
reply | reject}
<name>
AAA.ServerGroups.RADIUSServerGroupTable.RADIUSServerGroup.Authorization.Request aaa group server
radius
<group-name> {
authorization} {
accept | request }
<name>
AAA.ServerGroups.RADIUSServerGroupTable.RADIUSServerGroup.Accounting.Request aaa group server
radius
<group-name> {
accounting } {
accept | request}
<name>
AAA.ServerGroups.RADIUSServerGroupTable.RADIUSServerGroup.Accounting.Reply aaa group server
radius
<group-name> {
accounting } {
reply | reject}
<name>
4.3.x
OL-28375-03 275
AAA XML Support
XML CLI
AAA.ServerGroups.RADIUSServerGroupTable.RADIUSServerGroup.LoadBalance.Method.
LeastBounding
aaa group server
radius
<group-name>
load-balance
method
least-bounding
AAA.ServerGroups.RADIUSServerGroupTable.RADIUSServerGroup.SourceInterface aaa group server
radius group1
source-interface
AAA.ServerGroups.RADIUSServerGroupTable.RADIUSServerGroup.VRF aaa group server
radius
<radius-group>
vrf <>
AAA.ServerGroups.RADIUSServerGroupTable.RADIUSServerGroup.DeadTime aaa group server
radius
<radius-group>
deadtime <>
AAA.ServerGroups.RADIUSGroupTable.RADIUSGroup.PrivateServerTable.PrivateServer aaa group server
radius <>
server-private
<host>
RADIUS.Accounting show radius
accounting
RADIUS.Authentication show radius
authentication
RADIUS.Client show radius client
RADIUS.DynamicAuthorization show radius
dynamic-author
RADIUS.DeadCriteria.HostTable.Host show radius
dead-criteria host
<ip>
RADIUS.ServerGroups show radius
server-groups
DHCP XML Support
The support for XMLis available for DHCP that retrieves the client bindings, profile information, and DHCPv4
proxy statistics. It allows the management clients to perform client bindings based on Circuit-ID, Remote-ID,
Mac-Address, user profile information, and DHCPv4 proxy statistics. The mapping between CLI and XML
entries for the DHCP commands are as follows:
Release 4.3.x
276 OL-28375-03
DHCP XML Support
XML CLI
DHCPv4.ProfileTable.Profile.Proxy.RelayInformation.Check dhcp ipv4 profile
<name> proxy
relay information
check
DHCPv4.ProfileTable.Profile.Proxy.RelayInformation.AllowUntrusted
DHCPv4.ProfileTable.Profile.Proxy.RelayInformation.VPN
DHCPv4.ProfileTable.Profile.Proxy.RelayInformation.RemoteID
dhcp ipv4 profile
<name>proxy
relay information
option[vpn |
allow-untrusted |
remote-id
<name>]
DHCPv4.InterfaceTable.Interface.Proxy.Profile dhcp ipv4
interface
GigabitEthernet
<interface-name>
proxy profile
<name>
DHCPv4.ProfileTable.Profile.Proxy.RelayInformation.Policy dhcp ipv4 profile
<name>proxy
relay information
policy [drop | keep
| replace]
DHCPv4.ProfileTable.Profile.Proxy.VRFTable.VRF.HelperAddressTable.HelperAddress dhcp ipv4 profile
<name>proxy
helper-address [
vrf <name> ]
<server-ip-addr>
[ giaddr <ip-addr>
]
DHCPv4.ProfileTable.Profile.Proxy.BroadcastFlag.Policy dhcp ipv4 profile
<name> proxy
broadcast-flag
policy check
DHCPv4.ProfileTable.Profile.Proxy.ClassTable.Class
DHCPv4.ProfileTable.Profile.Proxy.ClassTable.Class.VRFTable.VRF.
HelperAddressTable.HelperAddress
DHCPv4.ProfileTable.Profile.Proxy.ClassTable.Class.Match.VRF
DHCPv4.ProfileTable.Profile.Proxy.ClassTable.Class.Match.Option
dhcp ipv4 profile
<name>proxy
class <class-name>
helper-address
[vrf <name>]
<server-ip-addr>
[ giaddr <ip-addr>
]
match vrf <name>
match option [ 124
| 125 | 60 | 77 ] hex
4.3.x
OL-28375-03 277
DHCP XML Support
XML CLI
<value> [ mask
<value> ]
DHCPv4.InterfaceTable.Interface.None dhcp ipv4
interface
<interface> none
DHCPv4.InterfaceTable.Interface.Proxy.CircuitID dhcp ipv4
interface
<interface> proxy
[information
optionformat-type
circuit-id <cir-id>]
DHCPv4.VRFTable.VRF dhcp ipv4 vrf
vrfname proxy
profile <name>
DHCPv4.NodeTable.Node.Proxy.Binding.ClientTable[DHCPv4ProxyCircuitIDFilter(Naming
CircuitID)]
show dhcp ipv4
proxy binding
circuit-id <cid>
location
<locationSpecifier>
DHCPv4.NodeTable.Node.Proxy.Binding.ClientTable[DHCPv4ProxyRemoteIDFilter(Naming
RemoteID)]
show dhcp ipv4
proxy binding
remote-id <rid>
location
<locationSpecifier>
DHCPv4.NodeTable.Node.Proxy.Binding.ClientTable[DHCPv4ProxyInterfaceFilter(Naming
InterfaceName)]
show dhcp ipv4
proxy binding
interface
<ifSpecifier>
DHCPv4.NodeTable.Node.Proxy.Binding.ClientTable[DHCPv4ProxyMACAddressFilter(Naming
MACAddress)]
show dhcp ipv4
proxy binding
mac-address
<addr> location
<locationSpecifier>
DHCPv4.NodeTable.Node.Proxy.Binding.ClientTable[DHCPv4ProxyBriefFilter] show dhcp ipv4
proxy binding
location
<locationSpecifier>
DHCPv4.NodeTable.Node.Proxy.Binding.ClientTable.Client show dhcp ipv4
proxy binding
detail location
<locationSpecifier>
DHCPv4.NodeTable.Node.Proxy.Binding.Summary show dhcp ipv4
proxy binding
Release 4.3.x
278 OL-28375-03
DHCP XML Support
XML CLI
summary location
<locationSpecifier>
DHCPv4.NodeTable.Node.Proxy.Binding.ClientTable[DHCPv4PProxyVRFFilter(Naming
VRFName)]
show dhcp ipv4
proxy binding vrf
<vrfname>
DHCPv4.NodeTable.Node.Proxy.ProfileTable.Profile show dhcp ipv4
proxy profile
name
<profile-name>
location
<locationSpecifier>
DHCPv4.NodeTable.Node.Proxy.VRFTable.VRF.Statistics show dhcp vrf
<name> ipv4
proxy statistics
location
<locationSpecifier>
DHCPv4.NodeTable.Node.Proxy.Statistics show dhcp ipv4
proxy statistics [
location < loc > ]
Control Policy XML Support
The support for XML is available for policy plane that retrieves subscriber management and subscriber session
related information. The mapping between CLI and XML entries for the control policy commands are as
follows:
XML CLI
InterfaceConfigurationTable.InterfaceConfiguration.ControlSubscriber.ServicePolicy interface
<intf>
service-policy
type control
subscriber
<policy-name>
Subscriber.Session.NodeTable.Node.SessionTable sh sub sess all
loc <loc>
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberDetailAllSessionFilter) sh sub sess all
detail loc
<loc>
Subscriber.Session.NodeTable.Node.Summary sh sub sess all
summary loc
<loc>
4.3.x
OL-28375-03 279
XML CLI
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberAllUsenameFilter) sh sub sess all
username loc
<loc>
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberInterfaceBriefFilter) {Naming
InterfaceName}
sh sub sess
filter
interface
<intf-name>
loc <loc>
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberInterfaceDetailFilter)
{Naming InterfaceName}
sh sub sess
filter
interface
<intf-name>
detail loc
<loc>
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberIPv4AddressVRFDetailFilter)
{Naming VRF Name, Address}
sh sub sess
filter
ipv4-address
<IPv4-addr>
loc <loc>
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberIPv4AddressVRFBriefFilter)
sh sub sess
filter
ipv4-address
<IPv4-addr>
detail loc
<loc>
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberMACAddressBriefFilter)
{Naming MACAddress}
sh sub sess
filter
mac-address
<mac-addr>
loc <loc>
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberMACAddressDetailFilter)
{Naming MACAddress}
sh sub sess
filter
mac-address
<mac-addr>
detail loc
<loc>
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberStateBriefFilter) {Naming
State}
sh sub sess
filter state
<state> loc
<loc>
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberStateDetailFilter) {Naming
State}
sh sub sess
filter state
<state> detail
loc <loc>
Release 4.3.x
280 OL-28375-03
XML CLI
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberUsernameBriefFilter) {Naming
Username}
sh sub sess
filter
username
<uname> loc
<loc>
Subscriber.Session.NodeTable.Node.SessionTable(SubscriberUsernameDetailFilter) {Naming
Username}
sh sub sess
filter
username
<uname>
detail loc
<loc>
sh sub sess
filter
ipv4-address
<IPv4 addr>
vrf <vrf> loc
<loc>
sh sub sess
filter
ipv4-address
<IPv4-addr>
vrf <vrf>
detail loc
<loc>
{Naming VRF Name, Address }
sh sub sess
filter vrf
<vrf-name>
loc <loc>
{Naming VRF Name, Address }
sh sub sess
filter vrf
<vrf-name>
detail loc
<loc>
Subscriber.Session.NodeTable.Node.SessionTable.Session{Naming SessionID} sh sub sess
sub-label
<0-ffffffff>
loc <loc>
Subscriber.Manager.NodeTable.Node.Statistics.AAA.Accounting sh sub man
stat AAA
accounting
loc <loc>
Subscriber.Manager.NodeTable.Node.Statistics.AAA.AggregateAccounting sh sub man
stat AAA
accounting
total loc <loc>
4.3.x
OL-28375-03 281
XML CLI
Subscriber.Manager.NodeTable.Node.Statistics.AAA.Authentication sh sub man
stat AAA
authentication
loc <loc>
Subscriber.Manager.NodeTable.Node.Statistics.AAA.AggregateAuthentication sh sub man
stat AAA
authentication
total loc <loc>
Subscriber.Manager.NodeTable.Node.Statistics.AAA.Authorization sh sub man
stat AAA
authorization
loc <loc>
Subscriber.Manager.NodeTable.Node.Statistics.AAA.AggregateAuthorization sh sub man
stat AAA
authorization
total loc <loc>
Subscriber.Manager.NodeTable.Node.Statistics.AAA.ChangeOfAuthorization sh sub man
stat AAA
COA loc
<loc>
Subscriber.Manager.NodeTable.Node.Statistics.AAA.AggregateChangeOfAuthorization sh sub man
stat AAA
COAtotal loc
<loc>
Subscriber.Manager.NodeTable.Node.Statistics.AAA sh sub man
stat AAA all
loc <loc>
Subscriber.Manager.NodeTable.Node.Statistics.AAA sh sub man
stat AAA all
total loc <loc>
Subscriber.Manager.NodeTable.Node.Statistics.AggregateSummary sh sub man
stats
summary
total <loc>
DAPS XML Support
The support for XML is available for distributed address pool service (DAPS) that retrieves the pool parameters
for distributed address pool services, and allows the management clients to get number of free, allocated and
excluded addresses based on VRF and pool name. The mapping between CLI and XML entries for the DAPS
commands are as follows:
Release 4.3.x
282 OL-28375-03
DAPS XML Support
XML CLI
PoolService.VRFTable.VRF.IPv4.Pool.Enable pool vrf <vrf-name> ipv4
<poolname>pool ipv4
<poolname>
PoolServce.VRFTable.VRF.IPv4.Pool.AddressRangeTable.AddressRange pool vrf <VRFName> ipv4
<PoolName> *
address-range <RangeStart>
<RangeEnd>pool ipv4
<PoolName> *
address-range <RangeStart>
<RangeEnd>
PoolService.VRFTable.VRF.IPv4.Pool.ExcludeTable.Exclude pool vrf <VRFName> ipv4
<PoolName> * exclude
<RangeStart>
<RangeEnd>pool vrf
<VRFName> ipv4
<RangeStart>
<RangeEnd>pool ipv4
<RangeStart> <RangeEnd>
PoolService.VRFTable.VRF.IPv4.Pool.UtilzationMark.High Pool vrf <VRFName> ipv4
<PoolName>
utilization-markhigh <>pool
ipv4 <PoolName>
utilization-mark high <>
PoolService.VRFTable.VRF.IPv4.Pool.UtilzationMark.Low Pool vrf <VRFName> ipv4
<PoolName>
utilization-mark low <>pool
ipv4 <PoolName>
utilization-mark low <>
PoolService.NodeTable.Node.VRFTable.VRF.IPv4 show pool vrf <vrf-name>
ipv4
PoolService.NodeTable.Node.PoolTable.Pool.IPv4.Detail show pool ipv4 name
<poolname>
PoolService.NodeTable.Node.PoolTable.Pool.IPv4.Verbose show pool ipv4 name
<poolname> verbose
PoolService.NodeTable.Node.VRFTable showpool ipv4showpool vrf
all ipv4
PPPoE XML Support
XML support is available for PPP over Ethernet (PPPoE) sessions. The mapping between CLI and XML
entries for the PPPoE feature commands are:
4.3.x
OL-28375-03 283
PPPoE XML Support
XML CLI
set PadoDelay.Default {<delay>} pado delay {<delay>}
set PadoDelay.CircuitId {<delay>} pado delay circuit-id {<delay>}
set PadoDelay.RemoteId {<delay>} pado delay remote-id {<delay>}
set PadoDelay.CircuitIdString{<string>} {<delay>} pado delay circuit-id string {<string>}
{<delay>}
set PadoDelay.CircuitIdSubString{<string>} {<delay>} pado delay circuit-id contains {<string>}
{<delay>}
set PadoDelay.RemoteIdString{<string>} {<delay>} pado delay remote-id string {<string>}
{<delay>}
set PadoDelay.RemoteIdSubString{<string>} {<delay>} pado delay remote-id contains {<string>}
{<delay>}
set PadoDelay.ServiceNameString{<string>} {<delay>} pado delay service-name string {<string>}
{<delay>}
set PadoDelay.ServiceNameSubString{<string>}
{<delay>}
pado delay service-name contains
{<string>} {<delay>}
set SessionIDSpaceFlat {TRUE} pppoe session-id space flat
PPPoECfg.BBAGroup {<group-name>} pppoe bba-group {<group-name>}
set PPPoE.EnableBBAGroup {<group-name>} pppoe enable bba-group {<group-name>}
set Tags.ACName {<name>} ac name {<name>}
set Tags.ServiceName(<name>).ServiceNameConfigured service name {<name>}
set Tags.ServiceSelectionDisable service selection disable
set Tags.PPPMaxPayloadDeny tag ppp-max-payload deny
set Tags.PPPMaxPayload {<min>,<max>} tag ppp-max-payload minimum {<min>}
maximum {<max>}
set MTU {<mtu>} mtu {<mtu>}
set Sesssions.MaxLimit {<limit>,<threshold>} sessions max limit {limit} threshold
{<threshold>}
set Sessions.AccessInterfaceLimit
{<count>,<threshold>}
sessions access-interface limit {<count>}
[threshold {<threshold>}]
set Sessions.MacLimit {<count>,<threshold>} sessions mac limit {<count>} [threshold
{<threshold>}]
set Sessions.MacIWFLimit {<count>,<threshold>} sessions mac-iwf limit {<count>} [threshold
{<threshold>}]
set Sessions.MacAccessInterfaceLimit
sessions mac access-interface limit
{<count>} [threshold {<threshold>}]
Release 4.3.x
284 OL-28375-03
PPPoE XML Support
XML CLI
set Sessions.MacIWFAccessInterfaceLimit
sessions mac-iwf access-interface limit
set Sessions.CircuitIDLimit {<count>,<threshold>} sessions circuit-id limit {<count>}
set Sessions.RemoteIDLimit {<count>,<threshold>} sessions remote-id limit {<count>}
set Sessions.CircuitIDAndRemoteIDLimit
{<count>,<threshold>,<radius-override>}
sessions circuit-id-and-remote-id limit
set Sessions.InnerVLANLimit {<count>,<threshold>} sessions inner-vlan limit {<count>}
set Sessions.MacThrottle
{<request-count>,<request-period>,<blocking-period>}
sessions mac throttle {<request-count>
<request-period> <blocking-period>}
set Sessions.MacAccessInterfaceThrottle
sessions mac access-interface throttle
{<request-count> <request-period>
<blocking-period>}
set Sessions.MacIWFAccessInterfaceThrottle
sessions mac-iwf access-interface throttle
<blocking-period>}
set Sessions.CircuitIDThrottle
sessions circuit-id throttle
<blocking-period>}
set Sessions.RemoteIDThrottle
sessions remote-id throttle
<blocking-period>}
set Sessions.CircuitIDAndRemoteIDThrottle
sessions circuit-id-and-remote-id throttle
<blocking-period>}
set Sessions.InnerVLANThrottle
sessions inner-vlan throttle
<blocking-period>}
set ControlPackets.Priority {<cos>} control-packets priority {<cos>}
set InvalidSessionID {DROP} invalid-session-id drop
set InvalidSessionID {LOG} invalid-session-id log
Subscriber Database XML Support
The support for XML is available for subscriber database that retrieves the subscriber association and session
information and allows the management clients to get subscriber session state, subscriber session information
based on unique subscriber label, subscriber association information based on unique subscriber label or
4.3.x
OL-28375-03 285
interface name or dynamic template name or type. The mapping between CLI and XML entries for the
subscriber database commands are as follows:
XML CLI
Subscriber.Database.NodeTable.Node.Association.LabelTable.Label{Naming
SubscriberLabel}
show subscriber database
association br location <>
Subscriber.Database.NodeTable.Node.Association.LabelTable.Label{Naming
SubscriberLabel}
association subscriber-label
<> br location<>
Subscriber.Database.NodeTable.Node.Association(SubscriberDatabaseLabelDetailFilter) show subscriber database
association location <>
Subscriber.Database.NodeTable.Node.Association(SubscriberDatabaseInterfaceBriefFilter){Naming
InterfaceName}
association interface-name
<> br location<>
Subscriber.Database.NodeTable.Node.Association(SubscriberDatabaseInterfaceFilter){Naming
InterfaceName}
<> location<>
Subscriber.Database.NodeTable.Node.Association(SubscriberDatabaseTemplateTypeBriefFilter){Naming
TemplateType}
association type <
ipsubscriber |ppp
|service-profile
|subscriber-service> br
location <>
Subscriber.Database.NodeTable.Node.Association(SubscriberDatabaseTemplateTypeFilter){Naming
TemplateType}
association type <
ipsubscriber |ppp
|service-profile
|subscriber-service>
location <>
Subscriber.Database.NodeTable.Node.Session(SubscriberDatabaseSessionStateFilter){Naming
Session-State}
session state <all| cfgapply
|cfgdone |cfggen
|cfgunapply |destroying
|error |fatgen |init |sync>
Subscriber.Database.NodeTable.Node.Session.LabelTable.Label{Naming
SubscriberLabel}
session subscriber-label <>
location <>
Subscriber.Database.NodeTable.Node.Association.LabelTable.Label association subscriber-label
<0x0-0xffffffff> brief
location R/S/M
<0x0-0xffffffff> brief
Release 4.3.x
286 OL-28375-03
XML CLI
<0x0-0xffffffff> location
R/S/M
<0x0-0xffffffff>
Subscriber.Database.NodeTable.Node.Association[SubscriberInterfaceBriefFilter(Naming
InterfaceName]
<ifname> brief location
R/S/M
Subscriber.Database.NodeTable.Node.Association[SubscriberInterfaceBriefFilter(Naming
InterfaceName]
<ifname> brief
Subscriber.Database.NodeTable.Node.Association[SubscriberInterfaceFilter(Naming
InterfaceName]
<ifname> location R/S/M
Subscriber.Database.NodeTable.Node.Association[SubscriberInterfaceFilter(Naming
InterfaceName]
<ifname>
Subscriber.Database.NodeTable.Node.Association[SubscriberTemplateType(Naming
TemplateType]
association type ppp brief
location R/S/M
TemplateType]
association type ppp brief
TemplateType]
association type ppp
location R/S/M
TemplateType]
association type ppp
TemplateType]
association type
ipsubscriber brief location
R/S/M
TemplateType]
association type
ipsubscriber brief
TemplateType]
association type
ipsubscriber location
R/S/M
TemplateType]
association type
ipsubscriber
TemplateType]
association type
subscriber-service brief
location R/S/M
TemplateType]
association type
subscriber-service brief
4.3.x
OL-28375-03 287
XML CLI
TemplateType]
association type
subscriber-service location
R/S/M
TemplateType]
association type
subscriber-service
TemplateType]
association type
service-profile brief
location R/S/M
TemplateType]
association type
service-profile brief
TemplateType]
association type
service-profile location
R/S/M
TemplateType]
association type
service-profile
TemplateType]
association type
user-profile brief location
R/S/M
TemplateType]
association type
user-profile brief
TemplateType]
association type
user-profile location R/S/M
TemplateType]
association type
user-profile
Subscriber.Database.NodeTable.Node.Association.LabelTable.Label association brief location
R/S/M
Subscriber.Database.NodeTable.Node.Association.LabelTable.Label association brief
Subscriber.Database.NodeTable.Node.Association.LabelTable.Label association location R/S/M
Subscriber.Database.NodeTable.Node.Association.LabelTable.Label association
Subscriber.Database.NodeTable.Node.Session.LabelTable.Label session subscriber-label
<0x0-0xffffffff> location
R/S/M
Subscriber.Database.NodeTable.Node.Session.LabelTable.Label session subscriber-label
<0x0-0xffffffff>
Subscriber.Database.NodeTable.Node.Session[SubscriberSessionStateFilter(Naming
State]
session state init location
R/S/M
State]
session state init
Release 4.3.x
288 OL-28375-03
XML CLI
State]
session state destroying
location R/S/M
State]
session state destroying
State]
session state cfggen location
R/S/M
State]
session state cfggen
State]
session state fatgen location
R/S/M
State]
session state fatgen
State]
session state cfgapply
location R/S/M
State]
session state cfgapply
State]
session state cfgdone
location R/S/M
State]
session state cfgdone
State]
session state cfgunapply
location R/S/M
State]
session state cfgunapply
State]
session state cfgerror
location R/S/M
State]
session state cfgerror
State]
session state error location
R/S/M
State]
session state error
State]
session state sync location
R/S/M
State]
session state sync
4.3.x
OL-28375-03 289
XML CLI
State]
session state all location
R/S/M
State]
session state all
Release 4.3.x
290 OL-28375-03
AP P E NDI X B
RADIUS Attributes
Remote Authentication Dial-In User Service (RADIUS) attributes are used to define specific authentication,
authorization, and accounting (AAA) elements in a user profile, which is stored on the RADIUS daemon.
This appendix describes the following types of RADIUS attributes supported in Broadband Network Gateway
(BNG):
RADIUS IETF Attributes, page 291
RADIUS Vendor-Specific Attributes, page 294
RADIUS ADSL Attributes, page 298
RADIUS ASCEND Attributes, page 299
RADIUS Microsoft Attributes, page 299
RADIUS Disconnect-Cause Attributes, page 299
RADIUS IETF Attributes
IETF Attributes Versus VSAs
RADIUS Internet Engineering Task Force (IETF) attributes are the original set of 255 standard attributes that
are used to communicate AAAinformation between a client and a server. Because IETF attributes are standard,
the attribute data is predefined and well known; thus all clients and servers who exchange AAA information
via IETF attributes must agree on attribute data such as the exact meaning of the attributes and the general
bounds of the values for each attribute.
RADIUS vendor-specific attributes (VSAs) derived from one IETF attribute-vendor-specific (attribute 26).
Attribute 26 allows a vendor to create an additional 255 attributes however they wish. That is, a vendor can
create an attribute that does not match the data of any IETF attribute and encapsulate it behind attribute 26;
thus, the newly created attribute is accepted if the user accepts attribute 26.
Table 7: Supported RADIUS IETF Attributes
Type Value Name
45 integer Acct-Authentic
4.3.x
OL-28375-03 291
Type Value Name
41 integer Acct-Delay-Time
52 integer Acct-Input-Giga-Words
42 integer Acct-Input-Octets
47 integer Acct-Input-Packets
85 integer Acct-Interim-Interval
51 integer Acct-Link-Count
53 integer Acct-Output-Giga-Words
43 integer Acct-Output-Octets
48 integer Acct-Output-Packets
46 integer Acct-Session-Time
40 integer Acct-Status-Type
49 integer Acct-Terminate-Cause
40 binary CHAP-Challenge
3 binary CHAP-Password
101 integer Dynamic-Author-Error-Cause
55 integer Event-Timestamp
11 binary Filter-Id
7 integer Framed-Protocol
8 ipv4addr Framed-IP-Address
22 "string" Framed-Route
14 ipv4addr login-ip-addr-host
50 string Multilink-Session-ID
32 string Nas-Identifier
4 ipv4addr NAS-IP-Address
5 integer NAS-Port
18 binary Reply-Message
6 integer Service-Type
32 string Tunnel-Assignment-Id
86 integer Tunnel-Packets-Lost
135 ipv4addr X-Ascend-Client-Primary-DNS
136 ipv4addr X-Ascend-Client-Secondary-DNS
Release 4.3.x
292 OL-28375-03
RADIUS Attributes
RADIUS IETF Attributes
Type Value Name
95 string NAS-IPv6-Address
123 binary Delegated-IPv6-Prefix
123 binary Stateful-IPv6-Address-Pool
97 binary Framed-IPv6-Prefix
96 binary Framed-Interface-Id
100 string Framed-IPv6-Pool
99 string Framed-IPv6-Route
98 string login-ip-addr-host
IETF Tagged Attributes on LAC
The IETF Tagged Attributes support on L2TP Access Concentrator (LAC) provides a means of grouping
tunnel attributes referring to the same tunnel in an Access-Accept packet sent from the RADIUS server to the
LAC. The Access-Accept packet can contain multiple instances of same RADIUS attributes, but with different
tags. The tagged attributes support ensures that all attributes pertaining to a given tunnel contain the same
value in their respective tag fields, and that each set includes an appropriately-valued instance of the
Tunnel-Preference attribute. This conforms to the tunnel attributes that are to be used in a multi-vendor network
environment, thereby eliminating interoperability issues among Network Access Servers (NASs) manufactured
by different vendors.
For details of RADIUS Attributes for Tunnel Protocol Support, refer RFC 2868.
These examples describe the format of IETF Tagged Attributes:
Tunnel-Type = :0:L2TP, Tunnel-Medium-Type = :0:IP, Tunnel-Server-Endpoint = :0:"1.1.1.1",
Tunnel-Assignment-Id = :0:"1", Tunnel-Preference = :0:1, Tunnel-Password = :0:"hello"
A tag value of 0 is used in the above example in the format of :0:, to group those attributes in the same packet
that refer to the same tunnel. Similar examples are:
4.3.x
OL-28375-03 293
RADIUS Attributes
IETF Tagged Attributes on LAC
Table 8: Supported IETF Tagged Attributes
Type Value IETF Tagged Attribute Name
64 integer Tunnel-Type
65 integer Tunnel-Medium-Type
66 string Tunnel-Client-Endpoint
67 string Tunnel-Server-Endpoint
69 string Tunnel-Password
82 string Tunnel-Assignment-ID
83 integer Tunnel-Preference
90 string Tunnel-Client-Auth-ID
91 string Tunnel-Server-Auth-ID
RADIUS Vendor-Specific Attributes
The Internet Engineering Task Force (IETF) draft standard specifies a method for communicating
vendor-specific information between the network access server and the RADIUS server by using the
vendor-specific attribute (attribute 26). Attribute 26 encapsulates vendor specific attributes, thereby, allowing
vendors to support their own extended attributes otherwise not suitable for general use.
The Cisco RADIUS implementation supports one vendor-specific option using the format recommended in
the specification. Cisco's vendor-ID is 9, and the supported option has vendor-type 1, which is named
"cisco-avpair." The value is a string of this format:
protocol : attribute sep value *
"Protocol" is a value of the Cisco "protocol" attribute for a particular type of authorization; protocols that can
be used include IP, IPX, VPDN, VOIP, SHELL, RSVP, SIP, AIRNET, OUTBOUND. "Attribute" and "value"
are an appropriate attribute-value (AV) pair defined in the Cisco TACACS+ specification, and "sep" is "="
for mandatory attributes and "*" for optional attributes. This allows the full set of features available for
TACACS+ authorization to also be used for RADIUS.
For example, the following AV pair causes Cisco's "multiple named ip address pools" feature to be activated
during IP authorization (during PPP's IPCP address assignment):
cisco-avpair= "ip:addr-pool=first"
If you insert an "*", the AV pair "ip:addr-pool=first" becomes optional. Note that any AV pair can be made
optional.
IETF Attribute 26 (Vendor-Specific) encapsulates vendor specific attributes, thereby, allowing vendors to
support their own extended attributes otherwise not suitable for general use.
cisco-avpair= "ip:addr-pool*first"
The following example shows how to cause a user logging in from a network access server to have immediate
access to EXEC commands:
cisco-avpair= "shell:priv-lvl=15"
Attribute 26 contains these three elements:
Release 4.3.x
294 OL-28375-03
RADIUS Attributes
Type
Length
String (also known as data)
Vendor-ID
Vendor-Type
Vendor-Length
Vendor-Data
It is up to the vendor to specify the format of their VSA. The Attribute-Specific field (also known as
Vendor-Data) is dependent on the vendor's definition of that attribute.
Note
Table 9: Supported Cisco Vendor-Specific RADIUS Attributes
Type Value Name
1 binary access-loop-encapsulation
1 string accounting-list
1 string acct-policy-in
1 string acct-policy-map
1 string acct-policy-out
1 integer actual-data-rate-downstream
1 integer actual-data-rate-upstream
1 integer actual-interleaving-delay-downstream
1 integer actual-interleaving-delay-upstream
1 integer attainable-data-rate-downstream
1 integer attainable-data-rate-upstream
1 string circuit-id-tag
2 string cisco-nas-port
1 string client-mac-address
1 string command
1 string connect-progress
1 integer connect-rx-speed
1 integer connect-tx-speed
1 string dhcp-client-id
4.3.x
OL-28375-03 295
RADIUS Attributes
Type Value Name
1 string dhcp-vendor-class
1 string disc-cause-ext
1 string disconnect-cause
1 integer if-handle
1 string inacl
1 boolean interworking-functionality-tag
1 string ip-addresses
1 string ipv4-unnumbered
This AVPair is preferred
for BNGin Cisco IOS XR
Software, and it is
equivalent to the
ip-unnumbered AVPair in
Cisco IOS Software.
Note
1 string login-ip-host
1 integer maximum-interleaving-delay-downstream
1 integer maximum-interleaving-delay-upstream
1 integer maximum-data-rate-downstream
1 integer maximum-data-rate-upstream
1 integer minimum-data-rate-downstream
1 integer minimum-data-rate-downstream-low-power
1 integer minimum-data-rate-upstream
1 integer minimum-data-rate-upstream-low-power
1 integer parent-if-handle
1 integer pppoe_session_id
1 string qos-policy-in
1 string qos-policy-out
1 string redirect-vrf
1 string remote-id-tag
1 string sa
1 string sd
1 string service-acct-list
1 string service-name
Release 4.3.x
296 OL-28375-03
RADIUS Attributes
Type Value Name
1 string sub-qos-policy-in
1 string sub-qos-policy-out
1 string traffic-class
1 string tunnel-tos-reflect
1 integer tunnel-tos-setting
1 string vpn-id
1 string vpn-vrf
1 integer vrf-id
1 string ipv6-addr-pool
1 integer ipv6-enable
1 integer ipv6-mtu
1 integer ipv6-strict-rpf
1 integer ipv6-unreachable
1 integer acct-input-gigawords-ipv6
1 integer acct-input-octets-ipv6
1 integer acct-input-packets-ipv6
1 integer acct-output-gigawords-ipv6
1 integer acct-output-octets-ipv6
1 integer acct-output-packets-ipv6
1 string delegated-ipv6-pool
1 string ipv6-dns-servers-addr
1 string dhcpv6-class
1 string ipv6_inacl
1 string ipv6_outacl
1 string addrv6
1 integer acct-input-gigawords-ipv4
1 integer acct-input-octets-ipv4
1 integer acct-input-packets-ipv4
1 integer acct-output-gigawords-ipv4
1 integer acct-output-octets-ipv4
1 integer acct-output-packets-ipv4
4.3.x
OL-28375-03 297
RADIUS Attributes
Vendor-Specific Attributes for Account Operations
Table 10: Supported Vendor-Specific Attributes for Account Operations
Action Type Value RADIUS AVP
account logon 1 string subscriber:command=account-logon
account logoff 1 string subscriber:command=account-logoff
account update 1 string subscriber:command=account-update
service activate 1 string subscriber:sa=<service-name>
service de-activate 1 string subscriber:sd=<service-name>
RADIUS ADSL Attributes
Table 11: Supported RADIUS ADSL Attributes
Type Value Name
144 binary Access-Loop-Encapsulation
142 integer Actual-Interleaving-Delay-Downstream
140 integer Actual-Interleaving-Delay-Upstream
130 integer Actual-Data-Rate-Downstream
129 integer Actual-Data-Rate-Upstream
134 integer Attainable-Data-Rate-Downstream
133 integer Attainable-Data-Rate-Upstream
1 string Agent-Circuit-Id
254 boolean social IWF-Session
141 integer Maximum-Interleaving-Delay-Downstream
139 integer Maximum-Interleaving-Delay-Upstream
136 integer Maximum-Data-Rate-Downstream
135 integer Maximum-Data-Rate-Upstream
132 integer Minimum-Data-Rate-Downstream
138 integer Minimum-Data-Rate-Downstream-Low-Power
Release 4.3.x
298 OL-28375-03
RADIUS Attributes
Vendor-Specific Attributes for Account Operations
Type Value Name
131 integer Minimum-Data-Rate-Upstream
137 integer Minimum-Data-Rate-Upstream-Low-Power
2 string Agent-Remote-Id
RADIUS ASCEND Attributes
Table 12: Supported RADIUS Ascend Attributes
Type Value Name
135 ipv4addr Ascend-Client-Primary-DNS
136 ipv4addr Ascend-Client-Secondary-DNS
196 integer Ascend-Connection-Progress
195 integer Ascend-Disconnect-Cause
187 integer Ascend-Multilink-Session-ID
188 integer Ascend-Num-In-Multilink
RADIUS Microsoft Attributes
Table 13: Supported RADIUS Microsoft Attributes
Type Value Name
30 ipv4addr MS-1st-NBNS-Server
31 ipv4addr MS-2nd-NBNS-Server
2 binary MS-CHAP-ERROR
28 ipv4addr MS-Primary-DNS
29 ipv4addr MS-Secondary-DNS
RADIUS Disconnect-Cause Attributes
Disconnect-cause attribute values specify the reason a connection was taken offline. The attribute values are
sent in Accounting request packets. These values are sent at the end of a session, even if the session fails to
4.3.x
OL-28375-03 299
RADIUS Attributes
RADIUS ASCEND Attributes
be authenticated. If the session is not authenticated, the attribute can cause stop records to be generated without
first generating start records.
lists the cause codes, values, and descriptions for the Disconnect-Cause (195) attribute.
The Disconnect-Cause is incremented by 1000 when it is used in RADIUS AVPairs; for example, disc-cause
4 becomes 1004.
Note
Table 14: Supported Disconnect-Cause Attributes
Description Value Cause Code
No reason is given for the
disconnect.
No-Reason 0
The event was not disconnected. No-Disconnect 1
Reason unknown. Unknown 2
The call has been disconnected. Call-Disconnect 3
Failure to authenticate number of
the calling-party.
CLID-Authentication-Failure 4
A modem in not available to
connect the call.
No-Modem-Available 9
No carrier detected.
Codes 10, 11, and 12 can
be sent if there is a
disconnection during
initial modemconnection.
Note
No-Carrier 10
Loss of carrier. Lost-Carrier 11
Failure to detect modem result
codes.
No-Detected-Result-Codes 12
User terminates a session.
Codes 20, 22, 23, 24, 25,
26, 27, and 28 apply to
EXEC sessions.
Note
User-Ends-Session 20
Timeout waiting for user input.
Codes 21, 100, 101, 102,
and 120 apply to all
session types.
Note
Idle-Timeout 21
Disconnect due to exiting Telnet
session.
Exit-Telnet-Session 22
Release 4.3.x
300 OL-28375-03
RADIUS Attributes
Could not switch to SLIP/PPP; the
remote end has no IP address.
No-Remote-IP-Addr 23
Disconnect due to exiting rawTCP. Exit-Raw-TCP 24
Bad passwords. Password-Fail 25
Raw TCP disabled. Raw-TCP-Disabled 26
Control-C detected. Control-C-Detected 27
EXEC process destroyed. EXEC-Process-Destroyed 28
User closes a virtual connection. Close-Virtual-Connection 29
Virual connected has ended. End-Virtual-Connection 30
User exists Rlogin. Exit-Rlogin 31
Invalid Rlogin option selected. Invalid-Rlogin-Option 32
Insufficient resources. Insufficient-Resources 33
PPP LCP negotiation timed out.
Codes 40 through 49
apply to PPP sessions.
Note
Timeout-PPP-LCP 40
PPP LCP negotiation failed. Failed-PPP-LCP-Negotiation 41
PPP PAP authentication failed. Failed-PPP-PAP-Auth-Fail 42
PPP CHAP authentication failed. Failed-PPP-CHAP-Auth 43
PPP remote authentication failed. Failed-PPP-Remote-Auth 44
PPP received a Terminate Request
from remote end.
PPP-Remote-Terminate 45
Upper layer requested that the
session be closed.
PPP-Closed-Event 46
PPP session closed because there
were no NCPs open.
NCP-Closed-PPP 47
PPP session closed because of an
MP error.
MP-Error-PPP 48
PPP session closed because
maximum channels were reached.
PPP-Maximum-Channels 49
Disconnect due to full terminal
server tables.
Tables-Full 50
4.3.x
OL-28375-03 301
RADIUS Attributes
Disconnect due to full internal
resources.
Resources-Full 51
IP address is not valid for Telnet
host.
Invalid-IP-Address 52
Hostname cannot be validated. Bad-Hostname 53
Port number is invalid or missing. Bad-Port 54
TCP connection has been reset.
Codes 60 through 67
apply to Telnet or raw
TCP sessions.
Note
Reset-TCP 60
TCP connection has been refused
by the host.
TCP-Connection-Refused 61
TCP connection has timed out. Timeout-TCP 62
TCP connection has been closed. Foreign-Host-Close-TCP 63
TCP network is unreachable. TCP-Network-Unreachable 64
TCP host is unreachable. TCP-Host-Unreachable 65
TCP network is unreachable for
administrative reasons.
TCP-Network-Admin
Unreachable
66
TCP port in unreachable. TCP-Port-Unreachable 67
Session timed out. Session-Timeout 100
Session failed for security reasons. Session-Failed-Security 101
Session terminated due to callback. Session-End-Callback 102
Call refused because the detected
protocol is disabled.
Invalid-Protocol 120
Disconnected by RADIUS request. RADIUS-Disconnect 150
Administrative disconnect. Local-Admin-Disconnect 151
Disconnected by SNMP request. SNMP-Disconnect 152
Allowed V.110 retries have been
exceeded.
V110-Retries 160
PPP authentication timed out. PPP-Authentication-Timeout 170
Disconnected by local hangup. Local-Hangup 180
Disconnected by remote end
hangup.
Remote-Hangup 185
Disconnected because T1 line was
quiesced.
T1-Quiesced 190
Release 4.3.x
302 OL-28375-03
RADIUS Attributes
Disconnected because the
maximum duration of the call was
exceeded.
Call-Duration 195
Call disconnected by client
(through PPP).
Code is sent if the LNS receives a
PPP terminate request from the
client.
VPN-User-Disconnect 600
Loss of carrier. This can be the
result of a physical line going dead.
Code is sent when a client is unable
to dial out using a dialer.
VPN-Carrier-Loss 601
No resources available to handle
the call.
Code is sent when the client is
unable to allocate memory (running
low on memory).
VPN-No-Resources 602
Bad L2TP or L2F control packets.
This code is sent when an invalid
control packet, such as missing
mandatory Attribute-Value pairs
(AVP), from the peer is received.
When using L2TP, the code will
be sent after six retransmits; when
using L2F, the number of
retransmits is user configurable.
VPN-Tunnel-Shut will be
sent if there are active
sessions in the tunnel.
Note
VPN-Bad-Control-Packet 603
Administrative disconnect. This
can be the result of a VPN soft
shutdown, which is when a client
reaches maximum session limit or
exceeds maximum hopcount.
Code is sent when a tunnel is
brought down by issuing the clear
vpdn tunnel command.
VPN-Admin-Disconnect 604
4.3.x
OL-28375-03 303
RADIUS Attributes
Tunnel teardown or tunnel setup
has failed.
Code is sent when there are active
sessions in a tunnel and the tunnel
goes down.
This code is not sent when
tunnel authentication fails.
Note
VPN-Tunnel-Shut 605
Call is disconnected by LNS PPP
module.
Code is sent when the LNS sends
a PPP terminate request to the
client. It indicates a normal PPP
disconnection initiated by the LNS.
VPN-Local-Disconnect 606
VPN soft shutdown is enabled.
Code is sent when a call has been
refused due to any of the soft
shutdown restrictions previously
mentioned.
VPN-Session-Limit 607
VPN call redirect is enabled. VPN-Call-Redirect 608
Release 4.3.x
304 OL-28375-03
RADIUS Attributes
AP P E NDI X C
Action Handlers
An action handler performs specific tasks in response to certain events. The following action handlers are
currently supported in BNG.
Authorization Action Handler, on page 305
Authentication Action Handler, on page 305
Disconnect Action Handler, on page 306
Activate Action Handler, on page 306
Deactivate Action Handler, on page 306
Set Timer and Stop Timer Action Handlers, on page 306
Authorization Action Handler
The authorization action handler obtains authorization data for a specific subscriber identity from external
AAA servers. The authorization action handler is an asynchronous function. It collects identity information
from Subscriber Attribute Database (SADB) as well as the user credential data based on the identifier type
specified in the CLI. This information along with method list name is sent to the AAA authorization
coordinator. Once the AAA processing is done, the control is returned to the Policy Rule Engine (PRE)
action handler to complete the event processing. The configuration example is as follows:
1 authorize aaa list <list-name> [identifier <identifier-type> | format <format_name>
password ['use-from-line'| <user-cfg-password>
Password is a mandatory, regardless of whether the user selects use-from-line or provides a specific value
to use for authorization.
Note
Authentication Action Handler
The authentication action handler gathers information like protocol type, service type, authentication type,
user name, chap attributes, and user password and passes them to the AAA coordinator along with the AAA
method list name. The authentication action handler is an asynchronous function. Once the AAA processing
is done, the control is returned to the PRE action handler to complete the event processing. The configuration
example is as follows:
1 authenticate aaa list <list-name>
4.3.x
OL-28375-03 305
Disconnect Action Handler
The disconnect action handler is called to disconnect a subscriber. For a subscriber disconnect, the PRE
informs the Policy Plane Session Manager (PPSM) to notify all clients about the subscriber disconnect. The
PPSM reports back to the PRE to complete the disconnection. The PRE puts the subscriber in the disconnect
state. The PRE also cleans-up the record history data that stores policy execution history and the control
block containing the subscriber label. When PRE processing is done, control is returned to the PPSM for
further processing.
Activate Action Handler
The activate action handler enables local dynamic templates or remote AAA services on the subscriber's
configuration. The results of this action are either immediate or asynchronous. The PRE gathers information
like the AAA method list name, template type, and template name and sends to the SVM for processing.
The SVM returns the control after completing template processing, and the PRE resumes processing the
action list from the place it had stopped. The configuration example is as follows:
1 activate dynamic-template <template-name> [aaa list <list-name>]
Deactivate Action Handler
The deactivate action handler disables local dynamic templates or remote AAAservices fromthe subscriber's
configuration. The result of this action is asynchronous. The PRE collects information like AAAlist, template
type, and template name and sends to the SVM. to request it to not apply the service. The AAA list is used
to derive a key used in SVM. SVMreturns control after completing template processing, and the PRE restarts
processing the action list from where it had stopped. The configuration example is as follows:
1 deactivate dynamic-template <template-name> [aaa list <list-name>]
Set Timer and Stop Timer Action Handlers
The set timer action handler sets an active named timer for a defined time period on the subscriber session.
The stop timer stops an active named timer on the subscriber session. Enabling the set timer action handler
allows the service provider to have one or more timed-policy-expiry events to be triggered on a subscriber.
This in turn provides better subscriber management over the subscriber life cycle. These action handlers
provide functions like scheduled validation of subscriber state status (checking if the subscriber is authenticated
or unauthenticated) and periodically changing subscriber policy (such as forcing re-authentication on a daily
or hourly basis).
An action with a timer value of 0, triggers the action immediately. Note
There are two methods to stop an active timer:
Allow the timer to expire.
Stop the active running timer using the stop-timer action command.
Release 4.3.x
306 OL-28375-03
Action Handlers
I NDE X
A
AAA 25
overview 25
aaa accounting command 30
aaa attribute format 33
aaa attribute format command 42
AAA attributes 31
aaa authentication command 30
aaa authorization command 30
aaa features 56, 63, 64
per-vrf aaa 63
radius change of authorization 56
radius double-dip feature 64
aaa group server radius command 37
aaa radius attribute command 39
access control list overview 215
access control lists 215, 217
configuring access-control lists 215
configuring access-groups 217
ipv4 access-groups commands 217
ipv4 access-lists commands 215
account-logoff 67
account-logon 67
Accounting 25
ACL 215
ACL and ABF support 215
acl-based forwarding overview 215
address/prefix pool 141
adjusting the tcp mss value of tcp packets 227
adsl attributes 298
ambiguous vlan support 159
ambiguous vlans 230
applying the policy to subscriber sessions 266
ASCEND attributes 299
ASR9001 22
Authentication 25
authentication disable command 113
authentication enable 113
authentication-no-response 67
Authorization 25
authorization-no-response 67
B
BNG 15, 16, 18, 21, 22
architecture 16
configuration process 21
hardware requirements 21
interoperability 22
overview 15
role in ISP network models 18
BNG packaging 19
C
caller id mask method remove match command 110
Calling-Station-ID 33
CGN 22
circuit-id 126
configuring a circuit-id for an interface 126
class class-default command 174
class-map 69
class-map type control subscriber match-any command 69
configuring a class-map 69
end-class-map command 69
match command 69
classification type 202
client lease time 128
CoA 56
conditional policers 202
configuration for ambiguous vlan 230
configuring a DHCP proxy 126
configuring a circuit-id for an interface 126
proxy information option format-type circuit-id command 126
configuring a policy-map 72
configuring a policy-map to run on the ipoe subscribers 87
configuring a policy-map to run on the pppoe subscribers 95
configuring aaa attribute format 42
4.3.x
OL-28375-03 IN-1
configuring aaa radius server 27, 30, 37, 39, 42
configuring aaa attribute format 42
configuring method lists for subscriber authentication 30
configuring RADIUS Attribute formatting 39
configuring RADIUS Attribute list 37
configuring RADIUS server group 27
configuring access-control lists 215
configuring access-groups 217
configuring aftr-name for ds-lite 165
configuring ambiguous vlans 159, 230
configuring control policies 69, 72
configuring DHCP proxy 124, 127, 130
attaching a proxy profile to an interface 130
configuring a remote-id 127
configuring DHCP ipv4 profile proxy class 124
dhcp ipv4 command 130
proxy class command 124
proxy profile command 130
relay information option remote-id command 127
configuring IGMP accounting 243
configuring igmp for subscriber interfaces 241
configuring ipoe sessions 82, 87
enabling ipv4 processing on an unnumbered interface 82
configuring ipv4 distributed address pool service 245
configuring ipv4 urpf 84
configuring ipv4 uRPF on subscriber interfaces 84
configuring ipv6 address or prefix pool name 162
configuring ipv6 ipoe subscriber interface 144
configuring ipv6 pppoe subscriber interfaces 152
configuring l2tp class options 113
configuring load balance for named server group 51
configuring load balancing for global server group 50
configuring logging for vpdn 109
configuring maximum simultaneous vpdn sessions 107
configuring method lists 30
configuring minimum bandwidth 235
configuring multicast 236, 239, 241, 243, 245
configuring IGMP accounting 243
configuring igmp for subscriber interfaces 241
configuring ipv4 distributed address pool service 245
configuring multicast hqos correlation mode 236
configuring route-policy for unicast qos shaper 239
configuring multicast hqos correlation mode 236
configuring options to apply on calling station id 110
configuring policy maps for http redirect 263
configuring policy on s-vlan 207
configuring pppoe sessions 74, 93, 94, 95
creating a ppp dynamic template with chap 94
creating a ppp dynamic template with pap 94
enabling a service-policy on an access interface 74
configuring pppoe sessions (continued)
enabling pppoe on access interface 93
configuring qos 190, 235
configuring minimum bandwidth 235
configuring qos accounting 190
configuring qos accounting 190
configuring RADIUS Attribute formatting 39
configuring RADIUS Attribute list 37
configuring RADIUS server group 27
configuring RADIUS server options 43
configuring route-policy for unicast qos shaper 239
configuring service accounting 60
configuring softshut for vpdn 115
configuring statistics IDs 62
configuring subscriber policy defined and applied through
RADIUS 174
configuring the class maps with or without redirection 261
configuring the client lease time 128
configuring the IGMP to HQoS correlation feature in a vrf 238
configuring the subnet number and mask for an address pool 248
configuring VLAN policy on an access interface 208
configuring vpdn template 105
configuring web logon with maximum time to authenticate 267
congestion control command 113
control policy 69, 72
control policy support 67
coupled policers 202
creating a configuration pool submode 246
creating a ppp dynamic template with chap 94
creating a ppp dynamic template with pap 94
D
DAPS server on RP 245
DAPS support 245
DAPS-Proxy on LC 245
defining vrf in a dynamic template 167
dhcp ipv4 command 130
DHCP IPv4 profile proxy 124
configuring DHCP IPv4 profile proxy class 124
DHCP proxy 127, 130
attaching a proxy profile to an interface 130
configuring a remote-id 127
dhcpv6 address/prefix pool 162
dhcpv6 ds-lite support 165
dhcpv6 features 143
dhcpv6 in global configuration mode 138
dhcpv6 in proxy profile configuration mode 138
dhcpv6 in server profile configuration mode 138
DHCPv6 overview 136
Release 4.3.x
IN-2 OL-28375-03
Index
dhcpv6 server and dhcpv6 relay or proxy 137
digest check disable command 113
Disabling Lawful Intercept 220
disconnect-cause attributes 299
distributed address pool service 245
dns-server 141
domain name 141
dynamic templates 75
E
enabling a service-policy on an access interface 74
enabling ipoe subscribers on an access interface 88
enabling ipv4 processing 82
enabling lac sso 102
enabling policy map merge 199
enabling pppoe on access interface 93
enabling vpdn nsr 104
Excessive Punt Flow Trap feature 211
F
FSOL handling 144
H
hardware requirements 21
hello-interval command 113
helper address 141
high availability in dhcpv6 143
hostname command 113
hqos correlation 234
http destinations with or without redirection 257
HTTP redirect feature 256
HTTP redirect statistics 256
I
ietf attributes 291
ietf tagged attributes 293
igmp accounting 243
igmp explicit-tracking command 241
igmp hqos correlation 202
igmp query-interval command 241
igmp query-max-response-time command 241
installing and activating the bng pie 19
ipoe sessions 87
event session-start match-all 87
ipoe sessions (continued)
policy-map type control subscriber command 87
ipv4 access-lists commands 215
ipv4 unnumbered command 82
ipv4 verify unicast source reachable-via rx command 84
ipv6 ipoe subscriber support 144
ipv6 pppoe subscriber support 152
L
l2 encapsulation 202
l2tp reassembly 101
l2tp session-id space hierarchical command 111
l2tp-class command 113
LAC SSO 102
lease 141
lease limit per circuit id 132
lease limit per interface 134
lease limit per remote id 133
logging command 109
lpts punt command 213
M
Merging QoS Policy-maps 198
method list 29
Microsoft attributes 299
modifying service policy through coa 187
multicast coexistence 233
multicast ipv4 command 236
multicast on pppoe interfaces 233
igmp accounting 233
multicast coexistence 233
multicast replication 233
multicast replication 234
multicast support 234, 243
igmp accounting 243
N
NAS-Port-ID 33
NAS-Port-Type 33
nas-port-type command 40
nV 22
4.3.x
OL-28375-03 IN-3
Index
O
open-garden 256
Option 82 Enhancements 136
Overview 25
Access, Accounting, and Authentication 25
P
pado delay command 117
parameterized subscriber policy 184
per-session lawful intercept 218, 219
per-vrf aaa 63
policy merging 202
policy-map 72
end-policy-map command 72
pool ipv4 command 245
ppp authentication chap command 94
ppp authentication pap command 94
PPP PTA 92
pppoe bba-group command 93
pppoe enable bba-group command 74, 97
pppoe session limit 119
pppoe session limit command 119
pppoe session throttle command 121
pppoe sessions 95
event session-start match-all 95
prefix delegation for dhcpv6 144
proxy class command 124
proxy information option format-type circuit-id command 126
proxy profile command 130
PTA 92
Q
QoS 173, 202
policy merging 202
QoS Accounting 189, 190
qos output minimum-bandwidth command 235
qos-account command 190
R
radius 50
load balancing 50
RADIUS 27, 29, 31, 291
attributes 31
method list 29
server group 27
radius adsl attributes 298
radius ascend attributes 299
radius attributes 291, 293, 294, 298, 299
radius adsl attributes 298
radius ascend attributes 299
radius disconnect-cause attributes 299
radius ietf attributes 291
radius Microsoft attributes 299
radius vendor-specific attributes 294
supported disconnect-cause attributes 299
supported ietf attributes 291
supported ietf tagged attributes 293
supported radius adsl attributes 298
supported radius ascend attributes 299
supported radius Microsoft attributes 299
supported vendor-specific radius attributes 294
radius change of authorization 56
radius CoA server 56
radius disconnect-cause attributes 299
radius double-dip feature 64
radius ietf attributes 291
radius load balancing 50
radius Microsoft attributes 299
radius vendor-specific attributes 294
radius-based lawful intercept 224
radius-server attribute command 27
radius-server throttle command 53
relay information option remote-id command 127
router igmp command 239, 243
S
server group 27
Service Accounting 58
service-policy input command 176
service-policy type control subscriber command 74
service-stop 67
session throttle 121
session-activate 67
session-limit command 107
session-start 67
SINT support 168
SNMP-based lawful intercept 218
softshut command 115
specifying a range of IPv6 addresses 250
specifying a set of addresses or prefixes inside a subnet 254
Specifying a Utilization Threshold 251
specifying the length of the prefix 253
Release 4.3.x
IN-4 OL-28375-03
Index
Statistics Infrastructure 62
subscriber session 79
support for shared policy instance 191
supported disconnect-cause attributes 299
supported ietf attributes 291
supported ietf tagged attributes 293
supported radius adsl attributes 298
supported radius ascend attributes 299
supported radius Microsoft attributes 299
supported vendor-specific radius attributes 294
T
TCP MSS Adjustment 226
template l2tp-class class_name command 105
template tunnel busy timeout timeout_value command 105
throttle command 55
throttling 53
timed-policy expiry 67
tunnel accounting command 113
type ppp command 241
U
understanding BNG 15
unicast-qos-adjust profile command 239
uRPFt 232
V
vendor-specific attributes 294
VLAN policy on access interface 206
vrf awareness 167
W
weighted queue limit 202
4.3.x
OL-28375-03 IN-5
Index
Release 4.3.x
IN-6 OL-28375-03
Index

B BNG Cg43xasr9k

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

B BNG Cg43xasr9k

Uploaded by

Copyright:

Available Formats

Cisco ASR 9000 Series Aggregation Services Router Broadband

Network Gateway Configuration Guide, Release 4.3.x

Cisco Systems, Inc. All rights reserved.

You might also like